Sanet - ST 1032301171

LEARNING AND TEACHING
FOR MATHEMATICAL
LITERACY
Typically, most people don’t realise when and how they can use the
mathematics they were taught in high school – yet many of the mathematical
ideas and skills can be a powerful tool for understanding how the world works.
Learning and Teaching for Mathematical Literacy addresses this situation,
offering practical strategies for developing a broader vision of mathematical
literacy in the classroom and recognising the importance of maintaining these
skills into adult life. Linked to the material explored throughout this book,
classroom activities and lesson materials are freely available for use via the
QR codes included in each chapter.
Filled with case studies and classroom activities, chapters tackle several
topics:
• Describing a framework for a broader vision of mathematical literacy –

what is it, and why is it important?
• Teaching mathematical literacy in the classroom.
• Applying mathematical literacy to ‘real-life’ scenarios: My dad is buying
a new dishwasher. Should he buy the extended warranty on offer? My
phone works fine, but I’ve been offered an upgrade. How should I decide
whether to take it?
• The role of technology in teaching mathematical literacy.
• Designing mathematical measures for real-world quantities.
Firmly grounded by practical applications for the classroom and

beyond, this is an essential handbook for any teacher, teaching assistant, or
mathematics subject lead who wishes to develop their students’ mathematical
literacy skills. This is also an ideal resource for those delivering or enrolled in
teacher preparation courses.
Hugh Burkhardt, long time Professor of Mathematical Education and Direc-
tor of Nottingham University’s internationally renowned Shell Centre, is an
applied mathematician and strategic researcher-designer in education. Work-
ing in both the United Kingdom and the United States, he is particularly
interested in the challenges of achieving change in school systems.
Daniel Pead has been an IT leader of the Shell Centre team since spending
his gap year working on the Investigations on Teaching with Microcomput-
ers as an Aid project in 1983. He has worked on the Centre’s contributions
to reSolve, Bowland Maths, and World Class Tests, amongst others. He is
particularly interested in the design of computer-based material for teaching
and assessment – preferably both together!
Kaye Stacey is an internationally renowned researcher, teacher educator, and

a prolific designer of educational materials. She is Emeritus Professor Math-
ematics Education at the University of Melbourne and chaired the Expert
Group for the 2012 PISA assessments of Mathematical Literacy.
All three authors have received awards for contributions to educational

design in STEM subjects from ISDDE, the International Society for Design
and Development in Education.
IMPACT (Interweaving Mathematics Pedagogy
and Content for Teaching)
The Learning and Teaching of Algebra

Ideas, Insights and Activities
Abraham Arcavi, Paul Drijvers, Kaye Stacey
The Learning and Teaching of Geometry in Secondary Schools

A Modeling Perspective
Pat Herbst, Taro Fujita, Stefan Halverscheid, Michael Weiss
The Learning and Teaching of Mathematical Modelling

Mogens Niss, Werner Blum
The Learning and Teaching of Number

Paths Less Travelled Through Well-Trodden Terrain
Rina Zazkis, John Mason, Igor’ Kontorovich
The Learning and Teaching of Calculus

Ideas, Insights and Activities
John Monaghan, Robert Ely, Márcia M.F. Pinto, Mike Thomas
The Learning and Teaching of Statistics and Probability

A Perspective Rooted in Quantitative Reasoning and Conceptual Coherence
Luis Saldanha, Neil J. Hatfield, Egan J Chernoff, Caterina Primi
For more information about this series, please visit: IMPACT: Interweaving Mathematics
Pedagogy and Content for Teaching - Book Series - Routledge & CRC Press
LEARNING AND
TEACHING FOR
MATHEMATICAL
LITERACY
Making Mathematics Useful
for Everyone
Hugh Burkhardt, Daniel Pead,

and Kaye Stacey
Designed cover image: © Shutterstock
First published 2024
by Routledge
4 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
605 Third Avenue, New York, NY 10158
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2024 Hugh Burkhardt, Daniel Pead and Kaye Stacey
The right of Hugh Burkhardt, Daniel Pead and Kaye Stacey to be identified as
authors of this work has been asserted in accordance with sections 77 and 78 of
the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or
utilised in any form or by any electronic, mechanical, or other means, now
known or hereafter invented, including photocopying and recording, or in any
information storage or retrieval system, without permission in writing from the
publishers.
Trademark notice: Product or corporate names may be trademarks or registered
trademarks, and are used only for identification and explanation without intent
to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
ISBN: 978-1-032-30116-7 (hbk)

ISBN: 978-1-032-30117-4 (pbk)
ISBN: 978-1-003-30350-3 (ebk)
DOI: 10.4324/9781003303503
Typeset in Sabon
by KnowledgeWorks Global Ltd.
CONTENTS
Series Foreword ix
Acknowledgements xi
Introduction 1
1 What Is Mathematical Literacy? 5
2 The Power of Tasks 30
3 Teaching for Mathematical Literacy 49
4 How Risky Is Life? 76
5 Climate Change: The Science, the Mathematics, and

the Politics 106
6 Planning for Good Things in Life 124
7 Looking Past the ‘Spin’ 140
8 Equality and Inequality 159
9 Your Money in Your Life 170

viii Contents
10 Computers in Teaching for Mathematical Literacy 191
11 The Importance of Curiosity 213
12 Designing Measures 226
13 Mathematics for Information Technology 240
14 Reflections 268
Index 274
Supplementary
material for this book and up-to-date links to
the online resources cited can be found online by scanning this
QR code or visiting ltml.mathlit.org.
SERIES FOREWORD
IMPACT, an acronym for lnterweaving Mathematics Pedagogy and Content

for Teaching, is a series of textbooks dedicated to mathematics education
and suitable for teacher education. The leading principle of the series is the
integration of mathematics content with topics from research on mathemat-
ics learning and teaching. Elements from the history and the philosophy of
mathematics, as well as curricular issues, are integrated as appropriate.
In mathematics, there are many textbook series representing internation-
ally accepted canonical curricula, but such a series has so far been lacking in
mathematics education. It is the intention of IMPACT to fill this gap.
Most of the books in the series focus on fundamental conceptual un-
derstanding of the central ideas and relationships of different branches of
mathematics. These central ideas and relationships serve as organisers for the
structure of each book. This book is rather different – more outward-looking.
It is built around the extra power that mathematics can give to people in their
everyday lives, interweaving the pedagogy and content that teaching for the
development of mathematical literacy involves for teacher educators, teach-
ers and their students in school.
x Series foreword
Most students spend many hours a week for many years studying math-
ematics. Historically, it has been the gateway to well-paid jobs – bookkeepers
used to need reliable skills in arithmetic while some professions, like engi-
neering, used algebra. But in this technological age, where those skills are
rarely needed outside the classroom, why does mathematics now have so
much more curriculum time than, say, Music – also an important and beauti-
ful aspect of human culture? What should school Mathematics now offer the
majority of students who are not going into STEM-heavy professions. This
book offers an answer, both theoretical and practical.
Series editors
Ghislaine Gueudet (France) Nathalie M. Sinclair (Canada)
and Günter Törner (Germany)
Series Advisory Board
Abraham Arcavi (Israel), Michèle Artigue (France), Jo Boaler (USA),
Hugh Burkhardt (Great Britain), Willi Dörfler (Austria),
Koeno Gravemeijer (The Netherlands), Angel Gutiérrez (Spain),
Gabriele Kaiser (Germany), Carolyn Kieran (Canada),
Frank K. Lester (USA), Fou-Lai Lin (Republic of China Taiwan),
John Monaghan (Great Britain/Norway), Mogens Niss (Denmark),
Alan H. Schoenfeld (USA), Peter Sullivan (Australia),
Michael 0. Thomas (New Zealand), Patrick W. Thompson (USA),
Francesca Ferrara (Italy), Eva Jablonka (Germany) and
Kyeong-Hwa Lee (South Korea)
Editorial Advisory Team for this book
Dr NG Kit Ee, Dawn (Singapore)
and Prof Dr Richard Barewell (Ottowa)
ACKNOWLEDGEMENTS
We would like to thank the Editors of the IMPACT series, Günter Törner,
Tommy Dreyfus, Ghislaine Gueudet, and Nathalie Sinclair for their invi-
tation to write this book. They and our families have provided continuing
support throughout. As to what you read, while the responsibility is ours,
it reflects many decades of working with wonderful colleagues around the
world, notably:
In the UK, the team at the Shell Centre for Mathematical Education of
the University of Nottingham, including (in alphabetical order) Alan Bell,
Barbara Binns, Jon Coupland, Rita Crust, Clare Dawson, Sheila Evans,
Colin Foster, Rosemary Fraser, John Gillespie, Steve Maddern, Andy Noyes,
Richard Phillips, John Pitts, Jim Ridgway and Geoff Wake, along with other
colleagues from a variety of collaborating institutions. David Spiegelhalter,
Daniel Burkhardt Cerigo, Laura Downton, Andy Jervis, Joe Fawcett and Les-
lie Dietiker helped with the chapters where their input is acknowledged.
Special mention must be made of Malcolm Swan, lead designer on so many
Shell Centre projects whose products are described here, and with whom one
of us (HB) shared the first Emma Castelnuovo Award of the International
Commission on Mathematical Instruction for “innovative, influential work
in the practice of mathematics education”.
In Australia Gary Asp, Terry Beeby, Lucy Bates, Lynda Ball, Jill Brown,
Susie Groves, Brian Low, John Malone, Katie Makar, Barry McCrae, Robyn
Pierce, Steve Thornton, Beth Price, Vicki Steinle, Gloria Stillman, Ross Turner,
David Leigh-Lancaster, Carly Sawatzki, Vern Treilibs, and Jill Vincent.
xii Acknowledgements
In the United States, Mary Bouck, Phil Daro, David Foster, Diane Schaefer,
Judah Schwartz, Ann Shannon, Sandy Wilcox and, especially, Alan Schoe-
nfeld with whom a 40-year collaboration in project leadership and writing
about insights gained has proved so valuable – and enjoyable.
Finally, thanks to the team at Routledge, particularly Bruce Roberts and
Lauren Redhead, who have produced this book so efficiently and expedi-
tiously. It has been a pleasure to work with them.
INTRODUCTION
Mathematical literacy is, roughly speaking, the ability to use your mathematics
in meaningful ways in a wide range of everyday life situations.
My dad is buying a new dishwasher. Should he buy the extended warranty

on offer?
Two friends and I are going on a vacation together this summer. How should
we set about planning it?
What is the chance that a teenager like my child will be in a shooting at
school? How does it compare to other hazards in their life?
These are typical of the challenges that people face in life. This book is about
how mathematics can help people to tackle them – and, more generally, to
better understand the world they live in and to make better decisions.
Mathematical literacy is now widely recognised as important for every-
one. In a world where technology does much of the technical ‘heavy lifting’
traditionally associated with mathematics and statistics, learning how to see
the essentials of life-related situations involves using mathematics, at what-
ever level, in new ways. In designing the school curriculum, the usefulness
of mathematics is a key justification for the large amount of time that all
children spend in Mathematics class.
In this book we look at the challenges and opportunities that this goal
presents for school mathematics and for teacher education. We aim to offer
direct support for:
• teacher educators, working in both initial teacher preparation and profes-

sional development contexts;
DOI: 10.4324/9781003303503-1
2 Introduction
• student teachers, especially those preparing to teach teenagers, and the

more experienced teachers they work with; and ultimately, through their
work,
• students in schools and colleges to whom those teachers teach mathematics.
The aim is to be readable by student teachers coming from a wide range

of mathematical backgrounds and to enable them to make mathematical
literacy part of their teaching. Our treatment of mathematical literacy encom-
passes other ‘literacies’ that feature in educational discussions – in particu-
lar statistical literacy, data literacy and aspects of computer literacy and the
mathematical sides of field-specific literacies such as scientific literacy, health
literacy, and financial literacy, not to mention language itself.
Because mathematical literacy, sometimes called ‘quantitative literacy’ or
‘numeracy’, is unlike most topics in the IMPACT series in being outside the
current practice of many mathematics teachers and most textbooks, this book
aims to be both:
• a descriptive and analytic text for reading about mathematical literacy –

intellectually sound yet lively, and
• a handbook that helps teachers to plan lessons in this new domain.
The form of support we have devised for these objectives is a combination

of description and analysis of the concepts involved together with direct help
with the challenges of turning these insights into classroom learning activi-
ties. To illuminate the descriptions and minimise misunderstanding, we use
a lot of examples – tasks and lessons that reflect practical situations where
using mathematics can have a payoff. We address the practical challenges
with examples of high-quality learning activities that address mathematical
literacy, with online links to the actual lesson materials.
Our focus will be on ways in which mathematics can give people more
power over real-world situations in the form of deeper understanding
through critical thinking about life-related contexts. We aim to develop a
‘big picture’ approach: using mathematics to help bring out the essentials
of a practical situation, to build insight into the situation, and support
good decisions without having to go into detail as a specialist might. We
explore, and illustrate through examples, everyday-life situations where
mathematics can be useful, noting the aspects of mathematics that play
central roles. We discuss ways in which the complementary foci of math-
ematical literacy and more traditional school mathematics can be used to
strengthen both.
Our aim throughout this book is to balance insight into the nature and
processes of mathematical literacy with a pragmatic focus on what will help
teachers to create and deliver effective and enjoyable mathematical literacy
Introduction 3
lessons in their classrooms. It is our hope that the reader, whether ‘leader’ or
‘student’, will find much that is useful, and even more that is interesting, in
the chapters that follow.
Structure of the book

The book is structured in three parts. In the first three chapters we set out our
approach to mathematical literacy. There follow six chapters that describe
and illustrate ways in which mathematics can inform understanding of spe-
cific areas that are important in everyday life. We then return to a deeper look
at aspects of mathematical literacy that have been just touched on before.
In the first theoretical chapter we describe and analyse the meaning of
mathematical literacy as set out, for example, by PISA. We present a Context-
Focused Mathematics Framework and illustrate it with complementary uses
of mathematics in modelling real-world situations and in getting meaning
out of data. The framework pays serious attention to the real-world context
at the centre, working alongside both knowing how to use mathematics and
knowing about its use. Additionally, the framework highlights productive
disposition to use mathematics and critical thinking at every stage. To com-
plement and enrich this description from a practical classroom perspective,
the next chapter analyses a varied range of exemplar tasks – some readers
may prefer to start here. The third chapter explores teaching mathematics
in a way that enhances mathematical literacy, and implications for the peda-
gogy and content of school mathematics curricula.
There follow six chapters on areas of practical importance discussed in
more detail: risk in everyday life; climate change, chosen for its far-reaching
consequences; planning for good things in life; looking behind the spin in
commerce and politics; the history and consequences of social equality and
inequality, and – the most well-trodden area – the roles of money in everyday
life. Our ‘big picture’ approach aims to bring out the essential features of
each context without too much detailed calculation. Mathematical literacy
is not about becoming an expert but, rather, being able to better understand
what experts do – and to formulate critical questions about their assump-
tions and the conclusions that follow. For example, we present and analyse a
lot of data; we encourage the reader to question the selections we have made
and the sources we list – a key aspect of being mathematically literate.
These chapters aim to show how mathematical literacy, as a widely ac-
cepted learning goal, can become part of the enacted mathematics curric-
ulum. To this end, we offer two kinds of support that session leaders, at
whatever level, may like to use to engage their students in active learning. We
suggest ‘activities’ for small group work on issues that arise in the text. We
outline lessons that have worked well in practice, many with links to well-
engineered lesson materials that are freely available online.
4 Introduction
The underlying messages in each of these six chapters on teaching for

mathematical literacy is the same – readers might not want to study all of
them. However, the chapters show something of the rich variety that mathe-
matical literacy offers for widening students’ understanding of the world out-
side the classroom. It thus makes mathematics part of the broader agenda of
critical thinking across school subjects in both the sciences and humanities.
The later chapters return to general issues of mathematical literacy. While
the use of computer technology is assumed throughout the book, we now
look in more depth at various specific ways it can enhance the learning of
mathematical literacy. We then turn to analyse curiosity– what it is and how
we may develop this prerequisite for wanting to think about life-related phe-
nomena from outside the classroom. As such it is an important and some-
times neglected part of the productive disposition that is identified in the
Context-Focused Mathematics Framework. We look at the process of de-
signing mathematical ‘measures’ of quantities of interest, examples of which
have arisen in all of the earlier chapters. We then look in more depth at the
mathematics of computer science that underpins digital technology from the
‘know about’ perspective that helps people understand the tools that they use
every day.
Finally, we return to reflect on what has been covered and how the teach-
ing of mathematical literacy fits into the roles that a teacher of mathematics
plays, in the classroom and beyond.
Links and references

We have included formal references to published books and papers in tra-
ditional format – as with all such citations, some of these may be available
to read online, others may require journal subscriptions or only be available
via academic libraries. In addition, we have included many links to resources
that are – at the time of writing – publicly available online. Some provide
reference material that may help with lesson ideas and planning, others offer
substantial collections of carefully designed teaching material. Rather than
print long web addresses, these links are in the form of codes like [1A] which
can be quickly looked up on the supporting website at ltml.mathlit.org. A
summary of links is included at the end of each chapter.
1
WHAT IS MATHEMATICAL LITERACY?
Mathematical literacy is the ability to use one’s mathematical capabilities to

better understand the world, its structures and events, and with that under-
standing to make better-informed decisions. In this chapter we explore the
meaning of mathematical literacy from a theoretical perspective, sketching its
history, the various ways it is described and defined, and its relationship to
the different aspects of school mathematics.
The term ‘mathematical literacy’ came to worldwide prominence at the
turn of the century with the advent of PISA, the widely influential OECD
Programme for International Student Assessment. PISA explicitly sets out to
assess mathematical literacy, scientific literacy, and reading literacy. The
PISA 2022 Mathematics Framework [1A] uses the following wording, setting
out clearly what is assessed:
Mathematical literacy is an individual’s capacity to reason mathematically

and to formulate, employ, and interpret mathematics to solve problems
in a variety of real-world contexts. It includes concepts, procedures, facts,
and tools to describe, explain, and predict phenomena. It helps individuals
know the role that mathematics plays in the world and make the well-
founded judgments and decisions needed by constructive, engaged and
reflective 21st Century citizens.
(Organisation for Economic Co-operation
and Development, 2013, p. 25)
We believe this broad description summarises the various ways in which

mathematics can be valuable to every citizen in their life beyond school, in-
cluding in the world of work. This is the focus of mathematical literacy in
DOI: 10.4324/9781003303503-2
6 What is mathematical literacy?
this book. Stacey and Turner (2015) discuss the issues involved in greater
depth than we have space for here.
Work on the development of mathematical literacy as a component of
school curricula has flourished for over half a century, much of it under
the broader umbrella of mathematical modelling (see e.g. Burkhardt, 2018;
Steen et al., 2007). Other terms have been used with much the same meaning.
Quantitative literacy is often preferred in the United States (Madison & Steen,
2008; Steen, 1999, 2002), whilst the term Numeracy was introduced by the
UK Crowther Committee in the 1950s (Crowther Report 15–18, 1959) as
“the mathematical equivalent of literacy”. The OECD’s Programme for the
International Assessment of Adult Competencies (PIAAC) [1B] sets out to
assess numeracy which it defines as “the ability to use, apply, interpret, and
communicate mathematical information and ideas”.
The overall goal of this chapter is to provide a rich description of mathe-
matical literacy as a basis for the discussion of learning and teaching mathe-
matical literacy in the next two chapters. There will be many examples from
various important context areas in the chapters that follow. In Section 1.1
we describe our context-focused mathematics framework that builds on
and integrates various theoretical approaches to mathematical literacy; this
theoretical framework links the diverse aspects we present in this book.
In Section 1.2 we explore the distinction between the ‘knowing how’ and
‘knowing about’ aspects of mathematics. In Section 1.3 we compare ‘theory-
driven’ and ‘data-driven’ approaches – constructing mathematical models of
practical situations and getting meaning out of data – and we look at these
two complementary perspectives in more depth in Sections 1.4 and 1.5.
These two skill sets feature throughout the book. In Section 1.6 we discuss
and illustrate the central role that technology plays in mathematical literacy,
throughout the book and in detail in Chapters 10 and 13. Section 1.7 draws
attention to the nomenclature problem: that mathematical literacy and even
more often numeracy are also used in a different, very narrow sense – as is
literacy itself.
1.1 A context-focused mathematics framework

There is a wide variety of theoretical structures that describe mathematical
literacy – what it is, or should be, and its place in the school curriculum, both
within mathematics and in other subjects. The review by Geiger et al. (2015)
provides an overview of these theoretical models, describing their varying
emphasis on ‘modelling with mathematics’ or ‘getting meaning out of data’.
Note that we use mathematics and mathematical literacy broadly, so it in-
cludes statistics, data literacy, and some aspects of digital literacy.
Building on these theoretical ideas, and our and others’ experience, we have
identified five core components of mathematical literacy, shown in Figure 1.1.1.
What Is mathematical literacy? 7
FIGURE 1.1.1 The context-focused mathematics framework.
We assert that mathematical literacy centres around taking the context seri-
ously, knowing about mathematical methods and ideas, and knowing how
to use some of them, and also draws on personal cognitive and attitudinal
characteristics – notably a critical thinking approach to situations and a dis-
position towards enquiry.
The five components of mathematical literacy
Taking the context seriously

In mathematical literacy, gaining more understanding of the life-related situ-
ation is the priority; in contrast, in school mathematics applications of math-
ematics are used mainly to exemplify and reinforce mathematical concepts
and techniques, which remain the focus. Solving mathematical literacy prob-
lems focuses on an understanding of the context that is deeper than in most
classroom applications of mathematics.
Knowing how to use a range of mathematical,

data, and technology skills
This is obviously central – always bearing in mind that individuals have dif-
ferent levels of, and needs for, mathematical expertise. The word ‘skill’ here
is used broadly to include concepts, processes, and procedures. Importantly,
as will become evident, important skills often differ from those most empha-
sised in traditional school mathematics. They reflect the widespread adoption
of information technologies in work and in everyday life, which has had a
profound effect on the mathematical knowledge needed for mathematical lit-
eracy. We include data to highlight not just the statistical knowledge of how
to process it but also aspects of its collection and quality.
Knowing about modelling, data, and technology

Whilst knowing how to create and adapt some mathematical models is
at the heart of mathematical literacy, the later chapters demonstrate that
a citizen in the modern world is confronted with many life-related con-
texts where the realistic models are highly complex, and well beyond the
capacity of non-experts to create. In these cases, ‘knowing about’ comes
to the fore. ‘Knowing about’ (understanding) the modelling process is
essential to interpret conclusions presented, in the light of the assump-
tions on which the model is based. Mathematical literacy also requires a
general appreciation of good practice in collecting, presenting, and inter-
preting data and a familiarity with the strengths and limitations of digital
technology.
Taking a ‘critical thinking’ approach

Focusing on the context and the limitations of mathematical models together
requires the questioning approach characteristic of critical thinking. This is
essential in order to solve life-related problems and to contribute productively
to discussing and evaluating the arguments of others. In school, mathematics
lessons are an opportunity to engage students in reasoning from evidence to
present and defend a case. A critical approach to data and its interpretation
is a key part of statistical literacy.
Demonstrating a productive disposition

A productive disposition requires confidence in applying mathematics
(for most people in most real-life situations, this mathematics will not
be complex), an expectation that using mathematics will be informa-
tive, and a willingness to use it. Productive disposition also requires a
willingness to persist through obstacles, to be prepared to think care-
fully and work flexibly, exercising initiative. Curiosity about phenom-
ena is an essential precursor of the kinds of thinking summarised above.
Whilst teachers can stimulate such thinking in the classroom, to become
mathematically literate citizens, students need to become self-motivated.
Mathematically literate individuals see mathematics as a useful tool for
exploring situations in which they are actively interested. Fostering this
may require encouragement to look beyond what is immediately relevant
to students today.
The rest of this chapter develops and illustrates these brief summaries.
This framework is an advance organiser, which will help to provide co-
herence across the diverse contexts that we discuss in the chapters that
follow.
1.2 ‘Know how’ versus ‘Know about’

We start with an important distinction that is not often explicitly set out.
Mathematics is unique among school subjects in that it is almost exclusively
a know how subject with very little know about. In schools, Mathematics
is focused very strongly on knowing how to perform procedures and solve
problems and the concepts that support this work; it generally includes very
little about the subject – its history, how concepts and ideas developed – and
even less about how mathematics beyond school level contributes to modern
life. Mathematical literacy does not always depend on a person knowing
how to solve a certain problem in all the detail that school mathematics typi-
cally requires. Instead, a mathematically literate person knows enough about
the concepts and techniques that experts employ to make well-informed and
reasonable decisions based on their advice. For example, a person using an
online app to see what the repayments would be for a loan does not need
to be able to do all the calculations themselves, but they do need to under-
stand that the app builds in assumptions about future interest rates, and need
to know enough about exponential behaviour to appreciate that apparently
small deviations in interest rates may alter repayments over time substan-
tially. There are many other examples throughout this book.
In summary, people thinking about challenging life-related problems need
to know about the context – the real-world situation of interest – and also
know about how mathematics (including statistics) is used by experts to help
bring out the essentials of what is going on. The balance between the two
will vary with the familiarity and complexity of the context. Because many
interesting and important life-related systems are often too complex for non-
experts to handle the analysis in detail, we aim to help readers to know about
and understand the principles of that expert-level analysis so that they can
form intelligent questions about the assumptions, methods, and conclusions
that people have offered. Even here a wide range of mathematical concepts
and skills are involved.
1.3 Complementary approaches – ‘theory-driven’

and ‘data-driven’
There are two complementary strands for approaching an interesting problem
situation, which we might call ‘theory-driven’ and ‘data-driven’. Sometimes
they are called ‘analytic’ and ‘descriptive’. In a ‘theory-driven’ approach, you
start by thinking about the situation and what quantities (variables) seem
important and how they may be related to each other (modelled) and what
you would expect (predict); this can then be tested with data and evaluated.
In the ‘data-driven’ approach, you start with some data that looks interesting
or is of concern, think what it might mean, and then consider how that phe-
nomenon might have arisen. Often, you use a mixture of these approaches,
sometimes starting with ideas about relationships, sometimes with data of

interest. For both approaches, you have to decide, at least, which variables
seem most important; in practice, your choice of variables is often made for
you by the data that is available and catches your attention – though, on
thinking the situation through critically, other data may be more informative
(see e.g. ‘excess deaths’ rather than ‘COVID deaths’ in Chapter 4).
Theory-driven approach – modelling a situation

The theory-driven approach to mathematical literacy focuses on using mathe-
matics to help describe and understand the structure of the problem situation.
The process is called mathematical modelling. There are many more-or-less-
equivalent descriptions of this process. One summary of the various phases
involved is shown in in Figure 1.3.1.
How does this process apply to tackling the Airplane Turnround task in
Figure 1.3.2?
When a typical class is presented with this task, their first response is to
just add up the individual times. That’s what you do in a mathematics les-
son. But once the teacher gets students to recognise that they must take the
context seriously – “Could you find a way to do it in less time?” – students
recognise that this is a different game and begin to think about the real-world
problem. Is there a way to do any of these things in parallel to reduce the
total turnround time? They then formulate a model by using their knowledge
of the turnround jobs to decide which must be sequential and which can be
done in parallel, going on to solve for the time for each parallel track by add-
ing the times of its components (a simple instance of solve). They interpret
that result by noting that the longest parallel track determines the turnround
time and, preferably in discussion with others, decide if their assumptions
are reasonable, calculations are correct, and if the result makes sense in the
FIGURE 1.3.1 The modelling process.

Source: After Burkhardt, 2008, and, National Governors Association, 2010.
Airplane turnround
Between landing and taking off, the following jobs need to be done on an
aircraft.
Job Time needed
A Get passengers out of the cabin and off the plane 10 minutes
B Clean the cabin 20 minutes
C Refuel the plane 40 minutes
D Unload the baggage from the cargo hold beneath the plane 25 minutes
E Get new passengers on the plane 25 minutes
F Load the new baggage into the cargo hold 35 minutes
G Do a final safety check before take-off 5 minutes
What is the shortest time needed to do all these jobs?
FIGURE 1.3.2 A task demonstrating taking the context seriously.

Source: Adapted from a task in Bowland Maths (see Section 6.1).
context – they evaluate the model. If it does not seem adequate, they go back
to formulate an improved model; if it looks good enough, they report the
result.
The phases in Figure 1.3.1 just exemplified for this mathematically simple
situation, can contain many hidden complexities. Note, in particular, how
the thinking has a mainly verbal form in the top half of Figure 1.3.1 and a
mathematical form in the lower part, moving back and forth between the
two as the modelling process proceeds.
Published variants of the modelling diagram have different emphases and
different amounts of detail. The above version emphasises the processes,
plunging down from the real world at the top into the mathematics below
and re-emerging with potentially useful answers. It makes explicit the itera-
tive nature of most modelling through the ‘decision point’ symbol Evaluate
and the important connections to the world of the initial problem through
the Problem and Report bubbles. We shall discuss modelling in more detail
in the next section.
As we shall see throughout the book, this theory-driven approach can use-
fully be applied to many life-related problems; models are rarely exact descrip-
tions of the situation but may still give useful insight – hence the aphorism “All
models are wrong but some are useful” (Box, 1979, pp. 202–203). Particularly
when the situation is complex, properly creating a theory-driven model may be
too difficult, yet modelling a few underlying essentials can still be informative.
Otherwise, one must turn to empirical data to gain some understanding.
Data-driven approach
The mathematical aspects of ‘getting meaning out of data’, sometimes called
data literacy, form the second major strand of this book. Interpreting data
presented in tables or graphs of various kinds is a core skill; deciding which
quantities are likely to be informative in developing understanding of the
situation and then finding that data is equally essential. The latter is usually
more challenging, although the web, search engines, and various technology-
enabled presentational techniques have made it both easier and more power-
ful. The results can be vivid, as the pioneering work of Hans Rosling and
collaborators with animated data illustrates – see Rosling and Rosling (2006),
Rosling (2010), Rosling et al. (2018) and the Gapminder website [1H]. We
shall discuss getting meaning out of data in more detail in Section 1.5.
Note that the data-driven approach tells you nothing about the mecha-
nisms that underlie the phenomena. For that you need to turn to model-
ling – mathematical or purely descriptive. This dialectic between these two
complementary approaches, data-driven and theory-driven, is at the heart of
understanding phenomena – and of mathematical literacy.
Critical thinking and critical enquiry

Critical thinking as outlined above is essential for all mathematical literacy.
However, it is especially key to approaches to theorising mathematical literacy
from the perspective of critical enquiry (see e.g. Jaworski, 2006). In this work,
contexts will generally be personal or societal, often with substantial political
implications that concern us as citizens. In this ‘critical thinking’ approach, the
role of mathematical knowledge is frequently mainly as a means of presenting
data, and providing support for asking critical questions – such as whether the
right variables have been selected, whether the methodology for data collection
and interpretation is sound, and what the real-world implications are.
The theoretical model of numeracy developed by Goos et al. (2019) em-
phasises a critical orientation – by which they mean “the use of mathematical
information to: make decisions and judgments, add support to arguments or
challenge an argument or position”. Like our framework, their model puts
contexts at the centre, and recognises the importance of mathematical knowl-
edge and a productive disposition to thinking mathematically, and they add
the importance of mathematical tools.
Command of language, more important in mathematics than is sometimes
recognised, is central to mathematical literacy, especially to achieving a criti-
cal orientation – for the analytic discussion of data, the issues it suggests, and
for the construction of powerful, logical arguments. Mathematics here may
be very much in a supportive role – but a potentially powerful one.
One strand of the diverse critical thinking community in mathematics
education (see e.g. Andersson & Barwell, 2021) focuses on global concerns,
some of which are the focus of later chapters here, with a socially proactive
approach that “is interdisciplinary; is politically active and engaged; is demo-
cratic; involves critique; and is reflexive and self-aware” (p. 7).
In the next two sections we examine the processes of mathematical model-
ling and of getting meaning out of data in a bit more depth.
1.4 Modelling real situations with mathematics

How do we set about using our mathematics to better understand a practi-
cal situation? This section develops the ‘theory-driven’ know how aspect of
mathematical literacy, particularly taking the context seriously, possessing a
range of appropriate mathematical and techno-mathematical skills, and be-
ing able to deploy them in understanding the situation. The long-established
term ‘mathematical modelling’ became part of everyday language only dur-
ing the 2020–2022 COVID-19 pandemic, when the predictions of the vari-
ous modelling groups were hot news. ‘Modelling’ has generally been thought
to need advanced mathematics; yet like Monsieur Jourdain, Moliere’s Bour-
geois Gentilhomme, who was surprised to learn that he had been speaking
prose all his life, most people have been modelling with mathematics at a ba-
sic level for many years – predicting the outcome of sharing sweets as young
children, working with money, planning an event.
The understanding of the modelling process and its phases, as depicted
in Figure 1.3.1, is not essential for using mathematics to tackle real-world
problems but is valuable for teachers in promoting awareness of the process
and for self-monitoring when tackling problems.
Various rather different activities are involved in modelling, moving from
the practical situation into a mathematical representation of some aspects
of it and back to the situation. Let’s talk it through using the task Sauce in
Figure 1.4.1 (assuming for the moment that it is a novel task rather than a
learned procedure for the solver). Even in a solution to this simple and rather
stylised task, the phases of modelling, as set out in Figure 1.4.2, are evident.
You are making your own batch of dressing for salads.

Here is a recipe for 100 millilitres (mL) of dressing Recipe for salad
taken from the web. dressing
Salad oil: 60 mL
Q1: How many millilitres (mL) of salad oil do you Vinegar: 30 ml
need to make 250 mL of this dressing? Soy sauce: 10 mL
FIGURE 1.4.1 Sauce.

Source: Adapted from PISA 2012 released mathematics item [1C].
The modelling process The outcome
It is important to get to understand both the A conceptual model of the

problem in its context and the task qualitatively. situation and the task.
What are the important quantities, the variables, This sauce is a mixture of
and how are they related. What is the question – three liquids and in this case
the task in hand? the volumes simply add (not
Read the Sauce task carefully, looking for the always true). The task is
significant quantities and the results asked for. to work out the scaled-up
(In designing assessment tasks like this PISA volumes of the ingredients,
mathematics task, every effort is made to make given the new total.
this phase easy.)
Next formulate a mathematical representation of A mathematical model of

the situation and the task. the situation and the task.
The variables are the volumes of each ingredient To change the total volume,
and the total in the recipe. The task is to represent all ingredient volumes are
the relationships when the total volume is changed multiplied by the scale
from 100 mL to 250 mL. factor, the ratio of the total
volumes.
Solve the problem mathematically by manipulating A mathematical result

the model to calculate the quantities the task 250/100 = 2.5
specifies. 60 × 2.5 = 150
Calculate the scale factor and carry out the 30 × 2.5 = 75
multiplication for each ingredient. Other 10 × 2.5 = 25
calculation methods are possible.
Interpret the mathematical result in the context of The result in context

the practical situation.
Link the mathematical answers to each ingredient. Oil 150 mL, vinegar 75 mL,
soy 25 mL
Evaluate the results in the context of the situation. These are either
Do they make sense? useful results or some
If so, use or communicate the result. insights to guide improving
the model.
If not, check the mathematics; if that seems correct,
go back to the beginning and improve the model
For Sauce the model seems to work well but we You can use these quantities,
have not looked at how well different quantities but if the desired volume
emulsify – but this probably would not change was much greater, you may
between 100 mL and 250 mL. need to mix them in small
quantities to ensure they
emulsify properly.
FIGURE 1.4.2 The phases of modelling illustrated by Sauce.

We shall have much more to say about modelling in action in later chap-
ters. A fuller discussion and analysis, including the development of these ideas
in various forms, can be found in Stacey and Turner (2015) and in another
book in this series: The Learning and Teaching of Mathematical Modelling
(Blum & Niss, 2020). Teaching materials focused on the modelling process can
be found in the reSolve Special Topic unit on Mathematical Modelling [1D]
(examples of which can be found in Chapters 4 and 10 in this book).
What is the value of this analysis of the modelling process, this ‘model of
modelling’, in developing mathematical literacy? In any process that requires
thought, it is usually helpful to be aware of what you are doing; particularly
when you get stuck and wondering what to do, such metacognitive reflection
helps to ‘break the log jam’.
• In formulating the mathematical model, usually the most challenging

phase, you may need to ask yourself questions like: Have I chosen the
right variables for what I’m trying to understand? What have I missed out
that is significant? Do I really need to consider this? How are these quanti-
ties related? How can I represent these relationships? Should I look at a
particular case in more detail? Would a spreadsheet help me to explore
possible ways forward in the light of the solutions that emerge? Do I need
further information? Some contexts are so important that students are
taught the appropriate models (e.g. to increase quantities in proportion,
relationships between distance, speed, and time, or for money conversion).
In others the relationships need to be constructed from first principles
when a new problem situation is tackled.
• The solve phase is well covered by school mathematics – though in non-
routine problem-solving, students can, in practice, only use concepts and
skills that they have thoroughly understood and connected to other parts of
mathematics and to a variety of contexts. Of course, it is also possible that
a solution to the mathematical problem is not available within the solver’s
level of mathematics; in which case, there is no alternative but to formulate
in a different, possibly simpler, way. Or perhaps numerical methods using a
graphing calculator or a spreadsheet might enable a solution. Many of the
complex models used by experts are only solvable in these ways.
• After the thinking done for the formulation phase, the interpret phase is
usually relatively straightforward, reflecting what was considered in for-
mulating the model. But the results may be surprising, with potentially
interesting life-related implications, or just wrong.
• For the evaluate phase, we are back in the real world, thinking about
the practical situation. Does the result make sense and does it provide a
useful answer to the question? If it looks dubious, check the mathematics
again. If it still doesn’t look right, do I need to formulate the problem in
another way?
We shall face questions like these in the examples throughout this book
and beyond. To summarise, the modelling diagram of Figure 1.3.1 can be
helpful both as a guide and as a checklist.
Finally, a word of caution. It is important to avoid unreasonable expec-
tations about the answers that a model can give – in particular, in predict-
ing the future. For example, modellers’ analyses of the COVID-19 pandemic
received a lot of publicity, with people expecting answers to ‘small’ details.
Should a lockdown limit people to 5 km or 10 km? Should people have to
wear a mask when outside? In fact, the model predictions were not ‘accu-
rate’; they were used by sensible people as the best guides available, month
by month, in a situation with many unknowns. The differences between dif-
ferent modelling groups gave some indication of the uncertainties. A model is
only as good as the assumptions and the data that go into it which, for real-
life systems, are rarely complete and accurate. Hence, again, “All models are
wrong but some are useful” – in itself inexact but useful guidance. (See, for
example, Pablo Rodríguez-Sánchez’s blog entry Can mathematical models
predict the future? [1E].)
Standard models versus active modelling

In taking the context seriously, there is a fundamental distinction that is im-
portant to bring out – between standard models or applications and active
modelling. When a new topic is introduced in school mathematics, it is often
accompanied by examples of that topic being used in various practical situa-
tions – for example, adapting a recipe as an application of proportional rea-
soning in Sauce, above. These standard applications are models of everyday
situations that the student is taught, primarily as examples of the mathemati-
cal topic in use in common situations of some importance. They serve:
• to reinforce understanding of the concept and develop skills through a

concrete embodiment,
• to equip students with models that describe some very common real-world
situations, along with
• the hope the student will recognise other situations where similar math-
ematical relationships are involved.
However, situations frequently arise in everyday life that have not been
covered in the classroom (or have been forgotten) but which people can
better understand by using their mathematics. Such situations require active
modelling. Figure 1.4.3 illustrates these very different mathematical activities.
Learning standard models and active modelling are complementary
activities – both are important in developing mathematical literacy. In teach-
ing standard applications, the emphasis is on students’ mastery of the new
FIGURE 1.4.3 Standard applications versus active modelling.
mathematical topic and canonical examples, which usually depict simplified

and stylised real-world situations. Tackling new situations, which demand
understanding the practical situation in some depth, requires active mod-
elling, using whatever mathematical tools seem likely to be useful in that
process. Learning standard applications is important in building a toolkit of
models that can be adapted to new situations as they arise. But recognising
possibilities in formulating a model is helped by learned connections to a
wider variety of situations than are usually brought in when the mathemati-
cal topic is the focus. Building these connections is at the heart of develop-
ing mathematical literacy. We can see this in Figure 1.4.4 from Be a Paper
Engineer – part of the Numeracy through Problem Solving materials dis-
cussed in Section 2.4.
FIGURE 1.4.4 Design and make a party invitation.

Source: Be a Paper Engineer – Shell Centre for Mathematical Education [1F].
To design the party invitation, formulation involves recognising that the

lengths of the horizontal and vertical sides of the pop-up are key variables,
and seeing their relationship to the pictures and the fold lines they show. For
the relationships, recognition of the length equalities is key – essentially the
parallelogram theorem on lengths. (Only a few students recognise it as such;
for others, the card brings the theorem to life.) Solving involves transferring
these insights in detail onto the net, using the pictures for scale and align-
ment. Interpreting and evaluation are best done by making the card.
Finally, a word of caution – again. The purpose of modelling is to give
some useful insights. But no model is perfect and many are far from it –
assumptions may be inadequate and/or significant variables may be ignored.
For example, money models often seem perfect because the whole situation is
a mathematical construct – the rules for adding interest to a savings account
are defined mathematically in a contract; but this model takes no account
of how the interest rate may be changed in the future. Moreover, it takes no
account of inflation which historically has meant that the real interest rate is
negative – you are actually paying people to borrow your money; estimated
corrections for that effect, called a ‘net present value’, are of course inexact.
In the end, the evaluation phase requires judgement based in real-world un-
derstanding and further critical reflection on what to do next. We look at this
and other aspects of financial literacy in Chapter 9.
This book is focused on modelling to better understand the real world,
with known mathematics. The Realistic Mathematics Education (RME)
project, developed at the Freudenthal Institute (see e.g. Van Den Heuvel-
Panhuizen, 2003), takes modelling as an approach to teaching and learn-
ing: starting with carefully chosen real-world or fantasy-world situations
from which mathematical concepts and representations can be abstracted
and developed. The two approaches are complementary and mutually
supportive.
1.5 Getting meaning out of data

The ‘data-driven’ aspect of mathematical literacy comes in when you are
presented with some data on a subject of interest or concern: starting from
this point is sometimes called data literacy or often, where random variation
is important, statistical literacy (Watson, 2006). For example, a report on a
bus crash on a school trip in which some students are killed might make a
teenager wonder about the chances of a fatal accident happening. Looking
for data on this, they might find Figure 1.5.1.
Understanding the information in the graphs will lead to some inferences.
The chance of a young person dying in the coming year is somewhere be-
tween 1 in 10 000 and 1 in 1 000. Even though the probabilities are very
low, from age 15 they are a lot higher for boys than for girls. This leads to
FIGURE 1.5.1 The average risk of dying in the next year.

Source: From Bowland Maths – How risky is life? [1J].
the immediate questions “Why?” then “What are the main causes of death,
and of the gender differences?” This leads onto the search for more data and,
sometimes, for models that explain it. We return to look at risk and the How
risky is life? materials in Chapter 4.
This is a typical example of the processes of data literacy. Starting from
a data-based mathematical description, in this case two graphs and their
axes, it resembles the interpretation and evaluation phases of the model-
ling process: understanding the meaning of information presented in various
forms – generally tables of numbers or graphs of various kinds – and seeing
how appropriate the data is to the situation of concern. Sometimes this leads
to trying to relate the data to underlying causes through seeking further data
or active modelling.
In a school context, this distinction is illustrated by the tasks Design and
Make a Party Invitation above, which involves the full modelling process, and
Hurdles Race (Figure 1.5.2) which just requires the interpretation and inter-
rogation of data presented as a graph. Hurdles Race (from Swan et al. 1985,
[1G]) is challenging in the following ways. The insights have to be assembled
into a coherent real-time narrative in the style of a radio commentary. Giving
a commentary from a graph like this is not going to arise in the real world,
but it is a valuable exercise in getting meaning out of graphs, a key part
of data literacy, as well as reinforcing the concepts involved. Hurdles Race
thus illustrates the mutual support that data interpretation and conceptual
The rough sketch shown above describes what happens when 3 athletes A, B
and C enter a 400 metres hurdles race.
Imagine that you are the race commentator. Describe what is happening as
carefully as you can. You do not need to measure anything accurately.
FIGURE 1.5.2 Hurdles race.

Source: From The Language of Functions and Graphs. Swan et al. 1985 [1G].
understanding can provide – a win-win for both mathematical literacy and

mathematical understanding (Figure 1.5.2).
Figure 1.5.3 gives several examples of data arising from real situa-
tions of wide interest which are discussed in later chapters of this book.
The first challenge is to interpret the graphs, identifying the variables and
scales and features of the data that are interesting. The next is to discuss
the implications – individual, societal, and political – of what is revealed.
The third is to begin to list further questions that arise and investigate where
data may be found to illuminate them. Finally, it is sometimes worth explor-
ing whether a mathematical model might be constructed that would offer an
analytic explanation and enable predictions.
The last 50 years have seen the development of very powerful ways, ini-
tially called ‘exploratory data analysis’ (Tukey, 1977), in which technology
can help us transform complex data into much more effective forms. Hans
Rosling (2010), in particular, developed ways of showing on screen many
kinds of historic population and human data and going beyond the tradi-
tional two variables by using colour and animation. That this work has im-
pacted mainstream media is a tribute to its power in getting meaning out of
data. In later chapters, we shall be concerned with data arising in a diverse
range of specific situations.
(a) Atmospheric CO2 levels since AD 1000. Data from NOAA – see Chapter 5.
(b) COVID-19 infection levels in England over time. Data from UK ONS – see
Chapter 4.4.
FIGURE 1.5.3 Data for classroom analysis and discussion. (Continued)

(c) UK gross domestic product 1992–2021. Data from World Bank – see
Chapter 7.3.
(d) Male versus female literacy (%) across countries. Data from UNESCO – see
Chapter 10.5.
FIGURE 1.5.3 (Continued)

1.6 The power of technology

In the world beyond school, most mathematical activity involves computer-
based platforms of various kinds – from calculators to cash registers in shops,
through mobile phones to supercomputers. That is how nearly all calculation,
information processing and data presentation is done nowadays. This reality
has been slow to penetrate school mathematics classrooms for multiple reasons.
For a long time, concerns about equality of provision between different schools
and different students led curriculum and examination authorities to exclude
technology. As ever, this influenced teachers and textbook writers. There is
also a view that progress in learning mathematics is covering more advanced
concepts but always exemplified only through simple tasks where technology
is less valuable. Further, curriculum time is always tight, so teaching students
to use technology can be seen as an extra burden. The situation has improved
steadily in this century, but technology-rich curricula remain the exception.
Most people still leave school with an acquaintance with a lot of math-
ematics that they cannot, in practice, apply to more than the simplest situa-
tions. Indeed, most adults use little if any of the mathematics that they first
meet after about age 14. The rapid, accurate execution of well-defined usu-
ally numerical procedures – which gained many people useful employment in
the distant past – is now the domain of technology. Yet there still seems to be
a lack of clarity over what needs by-hand mastery and what can be safely left
to technology. Mathematical literacy offers at least a partial justification for
the large amount of curriculum time that is still spent on mathematics but,
for this, it must be allied with technology. How?
The core of the approach is an alternative view of progress in mathemat-
ics – as the ability to tackle increasingly complex tasks using increasingly
sophisticated concepts and tools. (This is analogous to a widely accepted
definition of progress in first language studies – with tasks replacing texts.)
Task complexity – as opposed to the sophistication of the concepts – is the
new element. The real-world problems that mathematical literacy embraces
tend in that direction (as do those in really doing abstract mathematics – but
that’s beyond our brief here).
A simple example is solving quadratic equations. There is a famous for-
mula for the two solutions with its important and elegant proof, but too
many students forget it quickly and some never understand it. Teaching
about the quadratic formula introduces some widely applicable algebraic
techniques, but the formula itself only applies to quadratic equations. How-
ever, with easy calculation using technology, guess–check–improve tech-
niques are viable ways of solving equations – not elegant and not exact but
easy to carry out to as many decimal places as required, easy to remember,
and applicable to many types of equations, not just quadratics. Numerical
methods for solving equations with technology, refined to make them robust
and fast, are routinely used in industrial, engineering, and scientific work.
Mathematically literate citizens should at least know about this. Mathemat-

ics students should know how to use basic guess–check–improve methods
effectively and to use simple tools such as spreadsheets or calculators to
make them viable. Later chapters contain many examples of technology-
supported methods.
Simulation
Many systems are too complex to find mathematical relationships that de-
scribe them adequately, but they can be explored using computer-based
‘experiments’. In a simulation, analytic assumptions, including statistical
variability, are fed in at a ‘micro-level’ and subsequent macro-level behaviour
is calculated. This is repeated many times with minor changes to the initial
conditions and/or the parameters so as to understand the behaviour of the
system. Simulation can also be used in systems where, though the underlying
laws are completely understood and deterministic, the system behaviour is
very complicated or sensitive to minor changes in the initial conditions or
parameters. Simulation is now so important that it stands alongside theory
and experiment as one of the ‘pillars of science’.
Weather forecasting is a well-known example where large computer pro-
grams are used to simulate the atmosphere. Millions of individual observa-
tions of the atmosphere are obtained from satellites and a global network
of thousands of weather stations. The data is fed into the model, and the
future weather is simulated, running faster than it happens in reality. In these
systems, small variations in the data that are well within the uncertainties of
the observations can produce big changes in the predictions. This is fanci-
fully called ‘The Butterfly Effect’: a butterfly flaps its wings in one place and
produces a tornado far away. To make forecasts more reliable, weather fore-
casters run their calculations (simulations) many times with minor changes in
the input data. They then look for common features in the predicted weather
patterns – these features are used for the weather forecast. But it is expected
that occasionally one of the unusual patterns will actually happen.
1.7 The nomenclature problem

The book has a broad and ambitious interpretation of the term mathemati-
cal literacy. However, other narrower uses of the term are common and it is
important to be alert for the distinctions. Just as literacy is used in English
both for a broad knowledge of literary culture (being ‘well read’, etc.) and for
the narrow ability to read and write at a basic level (as opposed to illiterate),
both mathematical literacy and, particularly, numeracy are often narrowed
down to mean just competence in the basic skills of arithmetic. This ambigu-
ity has undermined the development of mathematical literacy in many educa-
tion systems. Even when mathematical literacy is an official goal, it is often
• A pair of jeans is on sale at “20% off” for $40. What was the original price?
• In a grocery store, 4 kg of tomatoes cost $5. How much will 7 kg cost?
• t seconds after an object falls from a height of H metres, its height h in
metres is given by the formula h = H − 5t 2
If H is 20 metres, when will the object hit the ground?
FIGURE 1.7.1 The narrow meaning for numeracy or mathematical literacy.
implemented narrowly as simply emphasising a subset of traditional school

mathematics, mainly arithmetic. This familiar material is easier to teach and
to assess than mathematical literacy as described here. But without the “capac-
ity to formulate, employ, and interpret mathematics in a variety of contexts”
– from the PISA definition – such skills are not enough to enable people to use
their mathematics to help with challenges of everyday life, or to ‘know about’
enough mathematics to satisfy their curiosity about other topics of interest.
Figure 1.7.1 exemplifies the narrow interpretation. All of the tasks there
might be described as “using mathematics to tackle a problem in everyday
life”, but they draw on only the most minimal understanding of the contexts.
The only discussion of context likely to happen is asking students who are
having difficulty with the third task: “What does ‘hit the ground’ mean about
the height?”
In contrast, the Airplane Turnround task in Figure 1.3.2 really does de-
mand thinking about the real-world situation. When students recognise that
the task requires considering which of the jobs can be done in parallel, they
can come up with a shorter time. There should be discussion about the vari-
ous assumptions being made – is it safe to refuel the plane when passengers
are on board? – and recognition that they would need more information to
make a fully informed decision. This is an example of “the sophisticated use
of elementary mathematics” (Steen, 2002; Steen & Forman, 2000) that dis-
tinguishes mathematical literacy from skill exclusively within mathematics.
The computational skill in this task is not demanding; the difficulty arises
instead from the need to work with both the situation and the mathematics –
critical thinking about the mathematical structure of turnround jobs done in
parallel or series, as well as the modest ‘technical skill’ of calculation.
That ambiguities of meaning arise is not surprising. (Mathematical lit-
eracy is also occasionally used in ways entirely unconnected with everyday
life – as a broad knowledge of pure mathematics in all its specialties, for
example.) The descriptive use of language is often ambiguous. Particularly
when something unfamiliar is described, people naturally interpret the words
within their prior experience. But there is a strategy that sharply narrows
such ambiguities: making meaning clearer through examples. We shall do so
throughout this book.
Using the term mathematical literacy in the broad sense of the PISA defini-
tion (and including, as noted above, statistical and other cognate literacies), the
obvious next question: “How broad?” The tasks in the next chapter, and in
those that follow begin to provide the answer. In presenting them, we will also
point to the essential roles played by the concepts, reasoning structures, and
the technical skills of abstract mathematics (by technical skills, we mean skill
in calculating, doing algebra, etc.). These are immediately clear to teachers in
the tasks in Figure 1.7.1; they are equally essential in tackling life-related tasks.
Another source of confusion when describing mathematical literacy is
that the versions that are valuable and feasible for different people differ im-
mensely in complexity. An adult working in a scientific or financial field will
need a much higher level of mathematical literacy than a young person just
starting out in the music industry might. We can say that both need math-
ematical literacy in their work and personal lives, but the level of mathemat-
ics used and the nature of the real-world situations they encounter are very
different. In this book, we focus on examples of mathematical literacy that
draw on typical school mathematics and involve life-related situations that
are likely to be of interest to many secondary school students in their current
lives and when looking forward. We also include some examples that arise in
more specialised work situations that are of wider significance – in finance,
for example.
For many problem situations, mathematical literacy involves this more
sophisticated use of simple-but-robust mathematics, particularly arithmetic,
proportional reasoning, and graphs. Along with resources that technology
offers, these provide the tools a mathematically literate person needs for both
simple modelling of practical situations with mathematics and ‘getting mean-
ing out of data’.
1.8 Summary and forward look

In this chapter, we have set out to clarify the meaning of mathematical liter-
acy through descriptions, definitions, and the context-focused mathematical
framework built from five broad components of mathematical literacy. We
have analysed the processes involved in modelling real situations with math-
ematics. We have begun to explore the processes of getting meaning out of
data. We have illustrated the power of technology and numerical methods in
both these areas. It is clear that mathematical literacy involves a deliberative
approach to thinking about a practical problem – what Daniel Kahneman
(2011) called “slow thinking” – critically questioning one’s instinctive, some-
times emotional initial reactions. The challenges and opportunities for teach-
ing that mathematical literacy presents are the focus of Chapter 3 – but first
we enrich the analysis of this chapter through a diverse range of examples of
classroom teaching for mathematical literacy.
Acknowledgements
Figure 1.4.1 is an adaptation of an original work by the OECD (PISA 2012
released mathematics item PM924Q02). The opinions expressed and argu-
ments employed are the sole responsibility of the author or authors of the
adaptations and should not be reported as representing the official views of
the OECD or of its member countries. PISA materials are licensed under the
Creative Commons Attribution-NonCommercial-ShareAlike 3.0 IGO (CC
BY-NC-SA 3.0 IGO) licence. Figures 1.4.4 and 1.5.2 are materials from the
Shell Centre for Mathematical Education and appear courtesy of the Bell
Burkhardt Daro Shell Centre Trust. Figures 1.3.2 and 1.5.1 appear courtesy
of the Bowland Maths maintainers.
References
Andersson, A., & Barwell, R. (2021). Applying critical mathematics education.
Brill.
Box, G. E. P. (1979). Robustness in the strategy of scientific model building. In R.
L. Launer & G. N. Wilkinson (Eds.), Robustness in statistics (pp. 201–236).
Academic Press. https://doi.org/10.1016/B978-0-12-438150-6.50018-2, ISBN
9781483263366.
Burkhardt, H. (2008). Quantitative literacy for all. In B. L. Madison & L. A. Steen (Eds.),
Calculation vs context: Quantitative literacy and its implications for teacher
education (pp 137–162). Mathematical Association of America. Retrieved May 19,
2023 from https://www.maa.org/sites/default/files/pdf/QL/cvc/CalcVsContext.pdf
Burkhardt, H. (2018). Ways to teach modelling: A 50 year study. ZDM, 50(1), 61–75.
https://doi.org/10.1007/s11858-017-0899-8.
Crowther Report 15–18. (1959). A report of the Central Advisory Council for Educa-
tion. HMSO.
Geiger, V., Goos, M., & Forgasz, H. (2015). A rich interpretation of numeracy for the
21st century: A survey of the state of the field. ZDM Mathematics Education, 47,
531–548. https://doi.org/10.1007/s11858-015-0708-1
Goos, M., Geiger, V., Dole, S., Forgasz, H., & Bennison, A. (2019). Numeracy across
the curriculum: Research-based strategies for enhancing teaching and learning
(1st ed.). Allen & Unwin.
Jaworski, B. (2006). Theory and practice in mathematics teaching development: Criti-
cal inquiry as a mode of learning in teaching. Journal of Mathematics Teacher
Education, 9(2), 187–211.
Kahneman, D. (2011). Thinking, fast and slow. Farrar, Straus and Giroux.
Madison, B. L. and Steen, L. A. (2008) Calculation vs context: Quantitative LIT-
ERACY and its implications for teacher education (pp. 137–162). Mathematical
Association of America. Downloaded as https://www.maa.org/sites/default/files/
pdf/QL/cvc/CalcVsContext.pdf
National Governors Association (2010) Common Core State Standards in Mathemat-
ics. Washington, DC
Niss, M., & Blum, W. (2020). The learning and teaching of mathematical modelling.
Routledge.
Organisation for Economic Co-operation and Development (OECD). (2013). PISA

2012 assessment and analytical framework: Mathematics, reading, science,
problem solving and financial literacy. OECD Publishing. https://doi.org/
10.1787/9789264190511-en
Rosling, H. (2010). Global population growth, box by box [Video]. TED Conferences.
https://www.ted.com/talks/hans_rosling_global_population_growth_box_
by_box
Rosling, H., & Rosling, O. (2006). How not to be ignorant about the world. Gap-
minder. Sweden. Retrieved May 19, 2023 from https://policycommons.net/arti-
facts/2475074/how-not-to-be-ignorant-about-the-world-gapminder/3497131/.
CID: 20.500.12592/s5h5hz.
Rosling, H., Rosling, O., & Rosling Rönnlund, A. (2018). Factfulness: Ten rea-
sons we’re wrong about the world: And why things are better than you think.
Flammarion.
Stacey, K., & Turner, R. (2015). Assessing mathematical literacy: The PISA experi-
ence. Springer.
Steen, L. A. (Ed.). (2002). Mathematics and democracy: The case for quantitative
literacy. National Council on Education and the Disciplines.
Steen, L. A. (1999). Numeracy: The new literacy for a data-drenched society.
Educational Leadership, 57(2), 8––13. Retrieved May 19, 2023, from https://www.
ascd.org/el/articles/numeracy-the-new-literacy-for-a-data-drenched-society
Steen, L. A., & Forman, S. L. (2000). Making authentic mathematics work for all
students. In A. Bessot & J. Ridgway (Eds.), Education for mathematics in the
workplace (pp. 115–126). Kluwer.
Steen, L. A., Turner, R., & Burkhardt, H. (2007). Developing mathematical literacy.
In W. Blum, P. L. Galbraith, H.-W. Henn, & M. Niss (Eds.), Modelling and appli-
cations in mathematics education. the 14th ICMI study. Springer.
Swan, M. with Pitts, J., Fraser, R., and Burkhardt, H, and the Shell Centre team
(1985). The language of functions and graphs. Shell Centre for Mathematical Edu-
cation. https://www.mathshell.com/materials.php?item=lfg&series=tss
Tukey, J. W. (1977). Exploratory Data Analysis. Adddison-Wesley.
Van Den Heuvel-Panhuizen, M. (2003) The didactical use of models in realistic math-
ematics education: An example from a longitudinal trajectory on percentage.
Educational Studies in Mathematics, 54, 9–35.
Watson, J. M. (2006). Statistical literacy at school: Growth and goals. Lawrence
Erlbaum.
Links to useful material

To visit any of these links, scan this QR code or visit ltml.math-
lit.org – append the link code to go directly to the entry – for
example, ltml.mathlit.org/1A
The original source links are given below for attribution
purposes:
[1A] PISA 2022 Mathematics Framework – OECD

https://pisa2022-maths.oecd.org/ca/index.html
[1B] Programme for the International Assessment of Adult Competencies

(PIAAC) – OECD
https://www.oecd.org/skills/piaac/piaacdesign/
[1C] PISA Released Mathematics Items 2012 – OECD
https://www.oecd.org/pisa/
https://www.oecd.org/pisa/pisaproducts/pisa2012-2006-rel-items-
maths-ENG.pdf
[1D] Special Topic: Mathematical Modelling – reSolve
https://www.resolve.edu.au/mathematical-modelling
[1E] Can Mathematical Models Predict the Future? – Pablo
Rodríguez-Sánchez
https://blog.esciencecenter.nl/can-mathematical-models-predict-the-
future-c362a0fbced2
[1F] Be a Paper Engineer – Shell Centre for Mathematical Education
https://www.mathshell.com/materials.php?series=numeracy&item=p
aperengineer
[1G] The Language of Functions and Graphs – Shell Centre for Mathemat-
ical Education
https://www.mathshell.com/materials.php?item=lfg&series=tss
[1H] Gapminder
https://www.gapminder.org/
[1J] How Risky Is Life? – Bowland Maths
https://www.bowlandmaths.org /projects/how_risky_is_life.html
2
THE POWER OF TASKS
We now move on from the general description and discussion of mathemati-

cal literacy to show something of the variety of tasks that learning mathe-
matical literacy involves, presenting and analysing task examples of different
kinds and lengths. Why the focus on tasks? We have seen in Chapter 1 how
illustrating descriptions with task examples clarifies the kind of thing one is
talking about. As we noted in Section 1.6, there is another, deeper reason to
give tasks a central role:
Progress in learning mathematics, and mathematical literacy, can be

thought of as the ability to tackle increasingly complex tasks using increas-
ingly sophisticated concepts and skills.
We aim to use the power of examples to build a down-to-earth ‘universe of

discourse’ about mathematical literacy, leading into a discussion of teaching,
pedagogy, and practice in Chapter 3.
All the tasks that follow work well in classrooms. Many have been
chosen because they come from teaching materials that have been imagi-
natively designed, carefully developed in classroom trials and are available
online. Such ‘well-engineered’ materials will be referenced throughout the
book, often with direct links, as resources for both teacher education and
school classrooms. The emphasis here is on ‘know how’, on students learn-
ing to solve life-related problems with mathematics, illustrating ways in
which the goals of developing mathematical literacy have been forwarded
by teachers.
DOI: 10.4324/9781003303503-3
The power of tasks 31
It is useful to identify some broad ‘task types’ that characterise mathe-

matical literacy and are shown in the examples that follow. Tasks may ask
students to do one of the following:
• Model and explain: invent, interpret, and explain models to understand real-
world phenomena and make predictions; modelling, as we have d escribed
in Chapter 1, is an important component of mathematical literacy.
• Plan and organise an event or a programme: find an optimum solution for
the goals of the event, subject to constraints of time, space, money, etc.
• Design and make an artefact or procedure, and test it: a common real-
world activity; the importance of thinking about imaginative design is often
overlooked.
• Explore and discover: finding relationships, often from looking at data in
some depth, leads on to hypotheses as to underlying causes, which may
later be expressed as models with predictions.
• Interpret, and translate between, representations: this is done with words,
numbers, graphs, or algebra to extract meaning from data, or from model
calculations.
• Evaluate and improve: review and improve an argument, a plan, or an
artefact. This kind of activity is less challenging than creating an argument,
plan, or artefact and can smooth the entry to a new context.
The aim of this list is not to give water-tight definitions that definitively
classify tasks (you might disagree with some of the classifications below), but
to help ensure that important kinds of mathematical literacy activity are not
overlooked in planning a mathematical literacy element in a curriculum. All
five components of the context-focused mathematics framework are impor-
tant in each of these areas.
The majority of tasks in this chapter require at least some parts of the
theory-driven mathematical modelling cycle. Later chapters will present
many data-driven problems. The mathematical techniques needed for the
tasks used here are standard (although not necessarily easy for students),
but each task also requires some strategic thinking especially related to the
formulating and evaluating phases of modelling discussed in Sections 1.3 and
1.4. This involves questions such as:
• Formulate
• How might I tackle this problem?
• How might I organise and represent the situation?
• Which bits of the mathematics I know could be useful?
• What assumptions do I need to make? What are the important variables?
• What are the mathematical relationships, if any, between the variables?
32 The power of tasks
• Evaluate
• Do my answers seem reasonable, given what I know of the real-world
situation?
• How can I justify my conclusion, explaining my reasoning?
This last point, reflected in the phrase “to describe, explain and predict
phenomena” from the PISA definition of mathematical literacy given in Chap-
ter 1, reflects the importance of students’ explanations and justifications – in
mathematical literacy and, equally, in really ‘doing mathematics’.
2.1 The range of tasks for mathematical literacy

We begin with the two simple tasks from PISA surveys in Figure 2.1.1. Both
could be used during the teaching of proportional reasoning.
Space flight is a model and explain task, built on recognising that propor-
tional reasoning applies – a scaling problem of a kind that arises regularly
in everyday life and work. Sauce in Figure 1.4.1 and the tomatoes task in
Figure 1.7.1 are other examples. Mathematically, there are several lines of
correct reasoning in ‘3-number proportion’ problems like this. Conceptually,
Space Flight
Space station Mir remained in orbit for 15 years and circled Earth some 86 500
times during its time in space.
The longest stay of one cosmonaut in the Mir was around 680 days.
Approximately how many times did this cosmonaut fly around Earth?
Ferris Wheel
A giant Ferris wheel is on the bank of
a river
See the diagram to the left
The Ferris wheel rotates at a constant
speed. The wheel makes one full
rotation in exactly 40 minutes
Q2: John starts his ride on the Ferris
wheel at the boarding point, P. Where
will John be after half an hour?
FIGURE 2.1.1 Two short mathematical literacy questions.

Source: PISA Released Mathematics Items [2A].
perhaps the most straightforward is to recognise a scaling situation, calculat-

ing the scale factor from the two given quantities that have the same dimen-
sions, in this case time. Converting both times to days, we get a scaling factor
to multiply by the total number of orbits, giving
680
× 86 500 ≈ 10 736
15 × 365.25
Ans: The cosmonaut flew around Earth about 11 000 times. The PISA
item indicated the accuracy required by the multiple-choice options sup-
plied: {110, 1100, 11 000, 110 000}. A mathematically literate person might
save calculation effort by estimating mentally: 90 000 orbits in 15 years is
6000 orbits per year, 680 days is about 2 years, so the answer is about
12 000 orbits. In a real situation, appreciating the accuracy that is pos-
sible from the data, and that is required to sensibly answer the question, is
critical.
Ferris Wheel (Q2), an interpreting task, is also about proportional reason-
ing, but part of the conceptual demand is visual and part dynamic – seeing
the movement in a circle over time, including the direction. Again, there are
various approaches. Probably the simplest is informal: to recognise that the
numbers (40 minutes and 30 minutes) are simply related, then to see that in
each 10-minute interval the wheel moves a quarter of a revolution, so point S
is the answer. The situation is an interesting one even if it is not obvious why
anyone would ask that question – perhaps when wondering where to look to
wave to one’s friends? One could generalise the argument, relating the angle
travelled to the time taken.
The set of released tasks from PISA [2A] is a far richer resource for teach-
ing mathematical literacy than these examples might suggest. It is a collection
every teacher of students in the PISA age group should explore. However, the
PISA task designers are constrained by the circumstances of an international
comparison survey. Each task must take most students no more than a few
minutes to complete, so that the complete survey can sample an adequate
range of real-world context types, mathematical content, and mathematical
processes. The task must also work in about 70 countries with their different
cultures and languages – ‘work’ in the sense of reliably allowing students to
understand the real-world situation, be likely to have learned the relevant
mathematics by age 15, and then to show what they can do. A challenging
brief indeed. The need for statistical robustness of the results provides further
constraints. These and many other issues are discussed in depth in Assessing
Mathematical Literacy (Stacey & Turner, 2015), their account of the design
and delivery of the PISA 2012 survey. The multiple-choice formats often
used in PISA, as in the original versions of these tasks, follow from these con-
straints. For classroom use, the tasks are usually best presented in an open
format, requiring a constructed response and explanation.
Upgrade? Table Tennis Tournament

My phone works fine but I've been Three of us have agreed to organise a
offered an upgrade. How should I table tennis tournament for our club.
decide whether to take it? How should we set about it? What
An evaluate and improve task information will we need?
A plan and organise task
FIGURE 2.1.2 Challenging situations for investigation.
Questions that arise in everyday life are usually broader and messier
than these tasks, so a mathematical literacy curriculum must include longer
and less-structured tasks that demand more extended chains of reasoning
than found in PISA, in typical classroom exercises, or in examinations. This
means that students should be expected to reason their way through the
multiple phases of modelling set out in Chapter 1: looking at what variables
are important, what assumptions to make about the relationships between
them, what data to collect to inform the analysis, and how to interpret and
evaluate the results – and to do all this without being led to a particular
solution by the teacher. Similarly, in a data-driven situation, students need
to interpret it not just by examining the data given and its statistical proper-
ties, but also considering its import in the context. In both cases, this also
means students having time to reflect on, discuss, and revise solutions. This
is all required for the two authentic tasks in Figure 2.1.2. Students will need
preparation to learn how to deal effectively with the demands of these very
open, extended tasks.
Figure 2.1.3 gives two examples of PISA tasks that can be used to help stu-
dents develop an investigative approach, understanding what is required and
developing confidence. Few students solved the Revolving Door task under
PISA test conditions, but it can work well in the classroom with teacher sup-
port through discussion. Although constrained, these tasks are more open
and involve more extended thinking than is usual. This will at first be a chal-
lenge for students, and a challenge for their teachers to support them, so the
examples that follow range from tasks only slightly longer and less structured
than typical test questions through to projects that can fuel several lessons’
work. The suggestion is to build into the curriculum a spectrum of tasks with
gradually increasing demand.
2.2 Sample 5–10-minute tasks

The tasks in Figures 2.2.1 and 2.2.2 (overleaf) each take about 5–10 minutes of
class time. Which Car is a PISA question where the main challenge is absorbing
and reviewing fairly complex data and then relating it to the specified constraints.
Continent area
Estimate the area of
Antarctica using the map
scale
Show your working out
and explain how you
made your estimate. (You
can draw over the map
if it helps you with your
estimation)
An interpret task
Revolving door
A revolving door includes three wings
which rotate within a circular-shaped space.
The inside diameter of this space is 2 metres
(200 centimetres). The three door wings
divide the space into three equal sectors.
The diagrams show the door wings in three
different positions viewed from the top.
Q2: The two door openings (the dotted
arcs in the diagram) are the same size. If
these openings are too wide, the revolving
wings cannot provide a sealed space and air
could then flow freely between the entrance
and the exit, causing unwanted heat loss
or gain. This is shown in the diagram
opposite.
What is the maximum arc length in
centimetres (cm) that each door opening
can have, so that air never flows freely
between the entrance and the exit?
FIGURE 2.1.3 Two richer tasks from PISA.

Which car?
Chris has just received her car driving licence and wants to buy her first car.
The table below shows the details of four cars she finds at a local car dealer.
Model: Alpha Bolte Castel Dezal
Year 2003 2000 2001 1999
Advertised price (zeds) 4800 4450 4250 3990
Distance travelled 105 000 115 000 128 000 109 000
(kilometres)
Engine capacity (litres) 1.79 1.796 1.82 1.783
Chris wants a car that meets all of these conditions:
The distance travelled is not higher than 120 000 kilometres.
It was made in the year 2000 or a later year.
The advertised price is not higher than 4500 zeds.
Q1: Which car meets Chris’s conditions?
FIGURE 2.2.1 An evaluate and improve task based on data.

Which Sport?
Which sport will produce a graph like this?
Choose the best answer from the following and explain exactly how it fits the graph.
Fishing
Pole Vaulting
100 metre Sprint
Sky Diving
Golf
Archery
Javelin Throwing
High Jumping
High Diving
Snooker
Drag Racing
Water Skiing
FIGURE 2.2.2 An interpret task.

Source: The Language of Functions and Graphs – Shell Centre for Mathematical Education [2B].
This is typical for tackling this kind of consumer choice decision. The straight-
forward strategies are to take the cars one by one checking the data against
each constraint, or to take the conditions one by one and eliminate cars at each
stage. Alpha is eliminated on price, Castel on distance, and Dezal on age. Bolte
scrapes through on all three which, in classroom use, could stimulate a critical
thinking discussion: “Is one year older serious when you could buy Dezal and
save 460 zeds, about 10% of the price of the car?” It is worth noting that the
engine capacity numbers are irrelevant to choosing on these criteria – sensible
since they are closely comparable; identifying which data is significant and
discarding redundant information is an essential part of mathematical literacy.
Which Sport? (Figure 2.2.2) is also about ‘getting meaning out of data’ –
here a speed-time graph. The presentation invites the common ‘picture-graph’
confusion, arising when students do not think carefully about the variables
on the axes. This will emerge and be sorted out in group and class discussion.
This task is from The Language of Functions and Graphs [2B] for which the
designer, Malcolm Swan, was awarded the prize for design excellence of the
International Society for Design and Development in Education (ISDDE).
‘LFG’ is an excellent teaching resource on the interpretation and sketching of
line graphs of real situations – a competence that is so important for getting
meaning out of data of many kinds.
Asking for permission (Figure 2.2.3) – a design and make task – explores the
other, human interaction, end of the demands of mathematical literacy. What in-
formation will the parent need to give informed permission for their child to go on
Asking for
permission
Imagine that your
class is going on
a trip to your
local swimming
baths next
Tuesday morning.
One person has
volunteered to
write a letter to
parents, asking for
permission.
Make a list of all
the important details
that have been
missed out of the
letter.
FIGURE 2.2.3 A design and make task.

Source: Plan a Trip – Shell Centre for Mathematical Education [2E].
this trip to the pool? How can it be presented in a convenient form that is likely
to get a positive response. This task requires logical, rather than mathematical,
thinking. It is included to show a short task that can scaffold students’ thinking and
build confidence to help them work an extended task, in this case the Plan a Trip
module (see Figure 2.4.1) from the Numeracy through Problem Solving series.
Students need time to consider each task carefully, and to work through it;
how much time depends on how far the teacher wants to take student expla-
nations and discussion – and consider the extensions that may emerge. A core
aspect of developing mathematical literacy in the classroom is tackling fewer,
more substantial tasks and thinking through the meaning of the mathematics
involved in greater depth. Slower can be quicker in learning.
2.3 Sample lesson-length tasks

Tasks that arise in everyday life often do not have ‘right answers’ – though they
nearly always have wrong ones! The tasks in this section come from the ‘Class-
room Challenges’ lessons of the Mathematics Assessment Project [2C]. These
materials provide substantial support for teachers. Many of the lessons focus
on concept development, designed to reinforce and deepen understanding of
curriculum topics, but about a third are problem-focused and highly relevant
to mathematical literacy. The project is described in Appendix 1 of this chapter.
Sharing gasoline costs (Figure 2.3.1) requires careful reasoning, with no
right answer. Students find issues of fairness important and motivating.
Sharing Gasoline Costs

Each day Lara’s mom drives her to school.
On the way, she picks up three of Lara’s
friends,
Chan, Jason and Marla.
Each afternoon, she returns by the same route
and drops them off at their homes.
This map is drawn to scale.
It shows where each person lives and the
route taken by Lara’s mom.
At the end of a term, the four students agree
to pay $300 in total towards the cost of the
gasoline.
How much should each person pay?

Try to find the fairest possible method. Show all your work.
FIGURE 2.3.1 A middle school evaluate and improve task.

Source: Mathematics Assessment Project Classroom Challenges [2C].
This leads to discussions in some depth on how to define fairness in this

context, each approach leading to a different algorithm to partition the cost.
For example, some students will charge in proportion to distance travelled;
others might charge less for blocks where there are more people in the car.
Others point out different aspects of the context that can tip the balance for
fairness – should savings from not catching the bus be considered? Sample
student solutions for discussion are given in the online lesson resources. An
evaluate and improve task where students compare various algorithms, this
is also the focus of a professional development module on Questioning and
reasoning from the Bowland Maths materials [2D].
Making fair divisions of cost is a rich topic for strengthening students’
critical thinking about the appropriateness of a direct proportion model in
different circumstances. As another example, how can two friends share the
cost fairly, if they buy two dresses together, using an offer of “buy one dress,
get 40% off a second (less expensive) dress”. Several potential models come
to mind: pay half the cost each, get half the saving each, divide the total cost
in proportion to the original prices. Making the fair decision involves details
of the social context as well as mathematical understanding of the conse-
quences of different procedures.
Making matchsticks [2C] (Figure 2.3.2), though the context is not directly
related to students’ lives, epitomises estimation problems and the mathemat-
ical challenges involved. The core concept here is ratios of volumes. Key
technical challenges are in handling big numbers, working to appropriate
Making Matchsticks
Matchsticks are 1
10 inch by 1
10 inch by 2 inches.
Matchsticks are often made from pine trees.
Estimate how many matchsticks can be made from
this tree:
80 feet tall
2 feet diameter at the base.
You may find some of the information given on the
formula sheet helpful.
Explain your work carefully, giving reasons for
any choices you make.
FIGURE 2.3.2 A high school task on volumes involving estimation.

Source: Mathematics Assessment Project Classroom Challenges [2C].
accuracy and dealing with the units correctly. Formulating the model involves
recognising that this can only be a rough estimate, cutting out the complexi-
ties of the tree and choosing a simple shape for the volume calculation: cyl-
inder or cone? A model and explain task, designed for the United States, the
non-decimal ‘customary units’ still in use there increase the challenge in some
ways but, because the factors 12 are explicit, errors made in the conversions
of length and volume stand out.
2.4 Sample multi-lesson tasks

Finally in this chapter, Figure 2.4.1 outlines examples of two mathemati-
cal literacy tasks that are of project length. Plan a Trip is a plan and
organise task from the UK Shell Centre Numeracy through Problem
Solving project [2E]. Similar Triangles (Ironing Table) is an explore and
discover task from the set of 16 substantial lessons for Years 7–10 from
the special topic Mechanical Linkages and Deductive Geometry by the
Australian reSolve Project [2F]. The following outlines of these two tasks
aim to help teachers decide whether to investigate these rich sets of online
teaching materials.
Numeracy through Problem Solving is a sequence of five modules that
develop students’ mathematical literacy. The five modules are Design

a Board Game, Produce a Quiz Show, Plan a Trip, Be a Paper Engineer,
and Be a Shrewd Chooser [2E]. They were designed by Malcolm Swan and
the Shell Centre team and developed with typical British teachers and their
mixed-ability classes. Designed for the age range 12–16, they have been used
Plan a Trip
In this module students plan and undertake a class trip out of school. This
involves costings, scheduling, surveys, and everyday arithmetic.
• In a card game simulation, groups undertake and record imaginary trips,
encounter problems and errors of judgment, then seek to correct them by
better planning.
• Students in a group share ideas for possible places to go and produce a leaflet
explaining these ideas. The class then work together to reach a decision on
the best destination and look at possible means of transport.
• The class lists and then shares out and undertakes the preparatory tasks that
need to be done before the trip can take place.
• The trip now takes place and, afterwards, the students reflect on what
happened, identifying successes and failures.
FIGURE 2.4.1 A plan and organise mathematical literacy project.

Source: Numeracy through Problem Solving – Shell Centre for Mathematical Education [2E].
successfully with younger children and with adults. Each module is designed
to take between 10 and 20 hours, usually over three weeks. The work is
primarily guided by a student booklet, with the teacher playing a monitoring
and ‘consultant’ role.
Each module works on a group-project basis and has four stages
(exemplified in Figure 2.4.1, for planning a day trip out of school). In Stage 1
students explore the context by working on and evaluating exemplars pro-
vided. Stage 2 is about generating and sifting ideas, which are developed
and implemented in detail in Stage 3. In Stage 4, each group evaluates the
designs that the other groups have produced. These are the essential phases
of most real-world problem-solving. A further example from this series, Pro-
duce a Quiz Show, is presented in Section 6.2. The Shell Centre Publications
website gives further information on the modules and access to teaching
materials.
Two kinds of assessment for these modules were provided. Formative
assessment, built into the teaching materials, is designed to check that
each student in a group understands all aspects of the work, not simply
those for which they may have been responsible. A final examination at
the end of the module assessed how well students could transfer what they
learned to more or less closely related problem situations. Administered
as formal e xaminations by the examination board, it showed one way
that mathematical literacy can be assessed in a high-stakes accountability
situation.
Mechanical Linkages and Deductive Geometry [2F], a reSolve: Mathemat-
ics by Inquiry special topic with four units aligned to the Years 7–10 cur-
riculum, shows students how geometry makes many things work. It provides
important applications of geometric theorems and stimulates the need for
proof by asking students to be sure that the things will always operate as
required. Mechanical linkages – sets of hinged rods – form the basis of many
everyday objects such as folding umbrellas, car jacks, scissors lifts, toolbox
lids, and pantographs. These objects work properly because of the geometry of
triangles, quadrilaterals, and circles. The lessons offer rich potential for group
exploration and justification. Students start by observing a real-life object,
identifying what must occur. For example, a car jack must rise vertically. They
then make a working model of the linkages from plastic strips or card and
paper fasteners and later use pre-prepared dynamic geometry software pro-
grams to investigate the underlying geometry. The software allows students
to observe more clearly what stays the same and what varies as the models
operate. The lessons are focused on angles and lines (Year 7), quadrilaterals
(Year 8), similar triangles (Year 9) with proof, and circle geometry for Year 10.
Extensive resources for teaching, including model templates, teaching notes,
and slides are provided (Figure 2.4.2).
Ironing Tables – from reSolve Mechanical Linkages (Similar Triangles unit)

When an ironing table with legs that pivot is raised or lowered, the top must
always stay parallel to the floor. How does this happen? Students investigate the
triangles formed by the pivoting legs, in particular investigating how different
leg lengths and pivot positions ensure that similar or congruent triangles are
formed. Three different designs can be considered, involving geometry of
different complexity.
• Students examine an ironing table or similar small folding table and identify
its important features, including size and shape for convenient ironing, need
for stability, adjustable height, and neat folding away. They may notice some
geometric features, such as the position of pivot point.
• Students create a physical model of the ironing table, observing more
geometric properties, especially as it moves.
• Using the dynamic geometry program, they measure angles and lengths,
and from this observe and later prove that two similar triangles are formed.
These make equal angles, so that the table is always parallel to the floor.
• Students are challenged to design an adjustable table with unequal legs.
FIGURE 2.4.2 An explore and discover lesson involving movement of everyday

objects.
Source: Mechanical Linkages and Deductive Geometry – reSolve [2F].
2.5 Task attributes

The examples of tasks in this chapter aim to illustrate something of the vari-
ety of ‘know how’ task types we will meet the chapters that follow. They give
some idea of what mathematical literacy, in all its variety, is about – as well
as pointing to some well-engineered resources for teachers to use in explor-
ing mathematical literacy in the classroom. We now draw these examples to-
gether by listing six characteristics that will help in selecting a collection that
represents a reasonably ‘balanced diet’ for developing the five components of
the context-focused mathematics framework (Section 1.1) across a diverse
range of contexts.
The type of task

Students should over time have experience of the six task types which began
this chapter:
• Model and explain real-world phenomena.

• Plan and organise events.
• Design and make an artefact, or design a procedure.
• Explore situations and discover relationships.
• Interpret representations in words, numbers, graphs, or algebra to extract
meaning.
• Evaluate and improve a solution to any of these.
The kind of context

PISA (see Stacey & Turner, 2015) divided the range of situations a task may
involve into four broad context categories:
• personal tasks of immediate relevance to the individual,

• societal tasks supporting informed citizenship,
• occupational tasks relevant to future employment, and
• scientific tasks (including science, engineering, economics, and pure
mathematics).
PISA includes situations from all of these categories to improve the validity
of its measures. Similarly, teachers can include all of these categories.
Most of the situations addressed in this book address personal and societal
contexts and concerns – the heart of mathematical literacy for most people –
though preparing for the other two kinds is very important. Especially for
young adults in vocational education, a focus on specific occupational tasks
can stimulate motivation for learning mathematics that they have not experi-
enced before. Mathematical literacy for scientific issues also has a workforce
benefit and often clear societal and personal implications – implementing health
advice sensibly, informing voting, and even knowing about the value of us-
ing logarithmic scales to report sound and earthquake intensity. Governments
invest in mathematics teaching because of the benefits of having a generally
mathematically literate workforce. There are mathematical literacy demands
specific to each job or profession, some basic and some at a very high level.
Authenticity
Traditional school mathematics contains many ‘real-world’ tasks, like those
in Figure 1.7.1, where the context is not intended to be seriously examined
by the students. But because mathematical literacy is characterised by taking
the context seriously, tasks used have a responsibility to be as authentic as

possible, given classroom constraints. This means they all need to be fairly
realistic with reasonable data, a focus on the big factors influencing the situa-
tion, expecting sensible solutions, and be clear whether the scenario is true or
invented. Answers need to be given to a sensible level of precision, reflecting
the precision in the data, assumptions made, and the demands of the con-
text. Tasks that foster mathematical literacy most strongly use contexts that
have an influence on the solution path. Of course, this does not exclude the
occasional fantasy situation or a classic problem that beautifully, if implau-
sibly, illustrates a mathematical point (such as the grains of wheat/rice on a
chessboard task [2G]. A mathematically literate person will appreciate how
extremely implausible the chessboard scenario is.)
Levels of expertise – ‘novice’, ‘apprentice’, and ‘expert’ tasks

A problem can be presented at various levels of sophistication to suit the level of
expertise of the solver. An ‘expert’ task is one presented as it might arise in life,
providing little or no guidance for how the problem is to be tackled, while an
‘apprentice’ version will break the problem down into several manageable steps.
A ‘novice’ task will either be an isolated exercise focused on a specific mathe-
matical procedure, or a collection of such tasks loosely grouped together under a
context (such as series of calculations that might be encountered ‘at the shops’).
Developing mathematical literacy is about improving ‘expertise’, as revealed in
responses to expert tasks; the other two types are important along the way.
Task difficulty
Traditionally, the difficulty of a mathematics task is judged primarily on the
technical level of the mathematics involved (e.g. on a scale from addition to
differential equations). However, the difficulty for students is also to affected
by the complexity and familiarity of the task situation, and the autonomy
expected of them in tackling the task. Unfamiliar tasks will be more difficult
than well-practised exercises, even where the underlying mathematics is the
same, as will tasks with more variables or more data, and those where the
student is expected to work out on their own how to approach the task, and
how to organise their work. So to keep tasks within the capacity of students,
less familiar and more complex tasks must be technically simpler – hence
Lynn Steen’s description of mathematical literacy as “the sophisticated use of
elementary mathematics” (Steen & Forman, 2000).
Openness
Many life-related problems don’t have a single formulation or a well-defined
correct answer, but allow for a diverse range of useful responses, possibly at
different levels of sophistication. Collections of tasks designed for mathemat-

ical literacy should include such open tasks and encourage critical thinking
to evaluate solutions from multiple viewpoints.
A more detailed discussion of the aspects of task design that foster mathe
matical literacy can be found in Chapter 14.
Appendix 1. The mathematics assessment

project’s Classroom Challenges
The Classroom Challenges [2C] were developed between 2010 and 2015
as part of a program to support the introduction of the US Common Core
State Standards in Mathematics. The aim was to produce teaching materi-
als which would encourage the formative assessment practices in Math-
ematics (Burkhardt & Schoenfeld, 2019) that had already proven highly
successful (Black & Wiliam, 1998). This model of lesson enrichment also
reflects Japanese approaches. About a third of these formative assessment
lessons focus on non-routine problem-solving, some of these in life-related
contexts.
The MAP team developed some design tactics that teachers may find
useful in promoting mathematical literacy, and elsewhere where a more
robust understanding of mathematics is desired. The top-level strategy is
to look in greater depth than usual at the task – of course, choosing tasks
that are rich enough to warrant that attention. To this end, the design of
the problem-solving lessons employs two particular design tactics that are
worth noting:
• A Common Issues table that lists difficulties that students showed in trials
of the materials, each linked to suggestions for non-directive interventions
(mostly questions) that a teacher might use to help the student move their
thinking forward and surface any underlying misconceptions that may be
blocking progress.
• Comparing student work where students are asked to compare carefully
constructed examples of ‘student work’ that use different methods, often
including numerical, graphical, and algebraic solutions. Besides making
the point that there are usually multiple valid ways to solve a problem, this
helps ensure that key elements of the curriculum are not lost and helps stu-
dents recognise why more sophisticated methods can be more powerful.
Each Classroom challenge begins with a 15–20-minute task that students

tackle individually in a prior lesson; this gives the teacher a chance to review
the responses, looking for different approaches and common issues. (Not
scoring, which is as counterproductive for formative assessment as it is time-
consuming and it distracts from misunderstandings.)
Developing your solution. The main lesson begins with any general com-
ments the teacher chooses to make, then 10 minutes or so for each student
to review and reconsider their ‘solution’, before moving to work with one
or two others for 20 minutes or so to produce a joint solution (felt-tipped
pens on poster-size paper allows the teacher to observe without ‘entering the
group’). The teacher observes, noting different approaches, only intervening
when a group is stuck – by choosing questions, perhaps from the Common
Issues table, that help the group think of alternative approaches, avoiding any
specific solution path. Groups then share their posters, encouraged to think
about questions as follows:
• Did they choose a good method for representing the situation?

• Did they make sensible assumptions?
• Is the reasoning correct? Are the calculations accurate?
• Are the conclusions sensible?
• Was the reasoning easy to understand and follow?
Collaborative analysis of the sample responses is the next main activity

(in the next class period if these are short). Each group discusses the three
or four sample responses provided, asking these questions and developing
written comments – assuming a ‘teacher role’. This activity gives students the
opportunity to see further approaches, and representations, and discuss the
assumptions made in each case. These are then shared in a whole-class dis-
cussion of the relative merits of various approaches. Finally, each group re-
turns to review its own solution and discuss possible changes it would make.
This outline will, as always, be brought to life by specific examples –
working through the lesson plan of one of these MAP lessons (see link [2C]),
for example:
Rolling cups – about the design of paper cups in the shape of a truncated
cone
or
Having kittens – modelling the number of descendants of an un-neutered
female cat
or
Muddying the waters – modelling the pollution of a lake as circumstances
change
The Math Assessment Project materials have seen considerable success,

with millions of downloads of individual Classroom Challenges modules.
Independent evidence shows substantial learning gains arising from their
use as part of a professional development program (CRESST, Research For
Action, 2015).
Acknowledgements
Extracts from The Language of Functions and Graphs (Figure 2.2.2),
Numeracy through Problem Solving, Plan a Trip (Figures 2.2.3 and 2.4.1)
are free for education use, and the Mathematics Assessment Project Class-
room Challenges (Figures 2.3.1 and 2.3.2) are available under the Creative
Commons CC BY-NC-ND 3.0 licence. Both of these appear here courtesy of
the Bell Burkhardt Daro Shell Centre Trust.
Several of the tasks referenced in this chapter are taken from the OECD’s
Program for International Student Assessment (PISA). These are adap-
tations of an original work by the OECD. The opinions expressed and
arguments employed are the sole responsibility of the author or authors
of the adaptations and should not be reported as representing the official
views of the OECD or of its member countries. The adapted PISA items
are licensed under the Creative Commons Attribution-NonCommercial-
ShareAlike 3.0 IGO (CC BY-NC-SA 3.0 IGO) licence. The specific PISA
items used in this chapter come from the Released Mathematics Items for
2006 and 2012:
Figure 2.1.1 Space flight – 2006 item M543Q01

Ferris wheel (Q2) – 2012 item PM934Q02
Figure 2.1.3 Continent area – 2006 item P01480
Revolving door – 2012 item PM995 (introduction and Q2)
Figure 2.2.1 Which car? – 2012 item M985Q01
Figure 2.4.2 is taken from reSolve: Mathematics by Inquiry © Australian

Academy of Science, available under Creative Commons CC-BY-NC-SA
licence.
See the Links section for online access to all of the above materials.
References
Black, P., & Wiliam, D. (1998). Inside the black box: Raising standards through
classroom assessment. King’s College.
Burkhardt, H., & Schoenfeld, A. H. (2019). Formative assessment in mathematics. In
R. Bennett, G. Cizek, & H. Andrade (Eds.), Handbook of formative assessment in
the disciplines (pp. 35–67). Routledge.
CRESST, Research for Action. (2015). MDC’s Influence on Teaching and Learning.
Research for Action. Retrieved from https://www.researchforaction.org/research-
resources/mdcs-influence-on-teaching-and-learning/ Accessed July 1, 2023.
Stacey, K., & Turner, R. (Eds.) (2015). Assessing mathematical literacy. Springer.
Steen, L. A., & Forman, S. L. (2000). Making authentic mathematics work for all
students. In A. Bessot & J. Ridgway (Eds.), Education for mathematics in the
workplace (pp. 115–126). Springer Netherlands.

To visit any of these links, scan this QR code or visit ltml.
mathlit.org – append the link code to go directly to the entry –
e.g. ltml.mathlit.org/2A
purposes.
[2A] PISA Released Mathematics Items 2012 and 2006 – OECD

https://www.oecd.org/pisa/
2006: https://www.oecd.org/pisa/testquestions-pisa2006.htm
2012: https://www.oecd.org/pisa/pisaproducts/pisa2012-2006-rel-
items-maths-ENG.pdf
[2B] The Language of Functions and Graphs – Shell Centre for Mathemat-
ical Education
https://www.mathshell.com/materials.php?item=lfg&series=tss
[2C] Classroom Challenges – Maths Assessment Project
Sharing gas: https://www.map.mathshell.org/lessons.php?unit=6200
&collection=8
Making matchsticks: https://www.map.mathshell.org/lessons.php?
unit=8300&collection=8
Rolling cups: https://www.map.mathshell.org/lessons.php?unit=9300
&collection=8
Having kittens: https://www.map.mathshell.org/lessons.php?unit=
9100&collection=8
Muddying the Waters: https://www.map.mathshell.org/lessons.php?unit=
9400&collection=8
[2D] Questioning and Reasoning – Bowland Maths
https://www.bowlandmaths.org.uk/pd/pd_05.html
[2E] Plan a Trip – Shell Centre for Mathematical Education
https://www.mathshell.com/materials.php?&series=numeracy
[2F] Mechanical Linkages and Deductive Geometry – reSolve
https://www.resolve.edu.au/mechanical-linkages
[2G] Wheat and Chessboard Problem
https://ltml.mathlit.org/2G
3
TEACHING FOR MATHEMATICAL
LITERACY
This chapter addresses the teaching of mathematical literacy – helping stu-

dents gain more power over the real-world situations they will encounter in
their present and future lives. Education in mathematics can empower stu-
dents to think more critically and constructively about their world.
3.1 The challenge

Mathematics, here including statistics, is a compulsory subject at school,
studied for multiple hours a week over many years. Why? Largely because
society as a whole appreciates that mathematics can make a unique contri-
bution to understanding practical situations and supporting good decision-
making in both personal and working lives. Why does mathematics have so
much more curriculum time than, say, music?
Having an adequate level of mathematical literacy is important for the life
chances of individuals. For the same reason, it is increasingly seen as having
national benefit in our interlinked, competitive, and digital world. Of course,
mathematics is not only taught for its utility: like music, it is also taught as
a significant part of cultural heritage with its elegant structures, intriguing
problem-solving opportunities, amazing theorems, and the special contribu-
tion it makes to logical thinking.
Both of these sides of teaching mathematics are important, but mathemati-
cal literacy is the essential goal for every student. If students leave school
without an understanding of how they can use the mathematics they have
learned, a general appreciation of how others use it, and a propensity (pro-
ductive disposition, critical thinking) for doing so when faced with chal-
lenges, then their mathematics education has not been really successful.
DOI: 10.4324/9781003303503-4
50 Teaching for mathematical literacy
This is increasingly recognised by countries around the world. It is a major part

of one of the United Nations Sustainable Development Goals for 2030 (n.d.),
where Target 4.6 is as follows:
By 2030, ensure that all youth and a substantial proportion of adults, both
men and women, achieve literacy and numeracy.
This represents a challenge for both teaching and teacher education in math-
ematics (Burkhardt, 2021; Madison & Steen, 2008). This chapter describes
ways in which this challenge can be more effectively met. There are three
main themes. The first is that developing mathematical literacy can be a part
of the majority of mathematics lessons from the beginning to the end of
education. Mathematical literacy can be fostered whenever teachers take the
time to link new mathematics to life-related situations. This involves teachers
taking the trouble to include problems and exercises with some degree of au-
thenticity, ensuring their students expect the mathematics to make sense and,
equally, that students think critically, using their common sense in discussing
how realistic the mathematical solutions are. However, it is also essential that
some lessons specifically target mathematical literacy, because of the different
priorities, summarized in Figure 1.4.3, when the focus is on understanding
the real-world situation. Then the mathematics, usually already well under-
stood, provides tools that help to illuminate the situation. This book gives
many examples of what these lessons might be.
The second theme, which will be addressed throughout this book, is that
mathematical literacy changes as the world changes, and change is especially
fast now as we are living in the digital revolution. The fluent pen-and-paper
mathematical skills that used to be essential for everyday life, or to get a
well-paying job, are rarely important now. Instead, a mathematically liter-
ate person is adept at using the mathematical tools that they have at hand,
from a simple calculator on a smartphone to a spreadsheet on a work or
home computer to programming very specialised software. School curricula
around the world are slowly adapting to this new situation but continuing
adaptation will be needed for decades to come. Technology advances on a
timescale much shorter than curricula can change. An important part of de-
veloping mathematical literacy is teaching students to use commonly avail-
able mathematical tools well. In the tasks presented throughout the book, we
assume students have access to appropriate digital tools.
The third theme underlying the chapter is the dependence of mathemati-
cal literacy on particular emphases in learning mathematics. Although every
piece of mathematical knowledge is potentially useful to illuminate some
life-related situation, some content is used very frequently, so deserves spe-
cial attention. Using mathematics also requires more reliable mathematics.
Beyond the classroom, there are likely to be many steps in solving real-world
Teaching for mathematical literacy 51
FIGURE 3.1.1 Context-focused mathematics framework.

Source: See Section 1.1.
problems, so students need robust skills and some autonomy in construct-

ing and explaining chains of reasoning, using problem-solving and model-
ling strategies, as well as a capacity for ‘self-monitoring’. These attributes
are summarised for mathematical literacy in the five key components of
our context-focused mathematics framework that we set out in Chapter 1
(Figure 3.1.1).
The implications for teaching will provide a connecting thread throughout
this chapter. Section 3.2 discusses how teaching for mathematical literacy
fits within the broader pedagogy of teaching mathematics, pointing to vari-
ations in teaching strategies that support mathematical literacy. In Sections
3.3 and 3.4 we look at the main topics that are taught in school curricula and
highlight some of the aspects of each topic that are especially important for
mathematical literacy. The idea is that developing mathematical literacy can
be part of almost every mathematics lesson. Section 3.5 discusses the value
of interdisciplinary activities, which give opportunities for students to engage
deeply with the context. Here we suggest that mathematics teachers some-
times work with teachers of other subjects so that the life-related problems
that are being studied can be treated with both mathematical depth and rich
understanding of the context. Section 3.6 looks at the ‘critical orientation’
that is part of mathematical literacy as a complex competency; in particular,
competency in examining evidence and creating and communicating argu-
ments and conclusions is recognised as part of the twenty-first-century skills
that are fundamental goals of education. Section 3.7 discusses what all math-
ematically literate students need to know about mathematical models – only
a few people will ever create significant mathematical models, but everyone
uses them, directly or indirectly, and needs the capacity to both appreciate
and critique them. Finally in Section 3.8 we discuss the professional devel-
opment for teachers that will help them help their students learn to become
mathematically literate.
3.2 Pedagogical variations for mathematical literacy

The fundamentals of teaching for mathematical literacy are no different from
those for teaching mathematics, and indeed other disciplines. But some dif-
ferences in emphasis arise from the central need to understand the real-world
phenomena involved.
The TRU dimensions

The essentials of fine mathematics teaching have been extensively dis-
cussed and researched over centuries – and increasingly probed in a wide
range of research programmes. Here we shall use the distillation of best
practice set out in the Teaching for Robust Understanding dimensions, de-
veloped by Alan Schoenfeld and teams of collaborators [3A]. TRU asserts
that the quality of a learning environment depends on the extent to which
it provides opportunities for students along the five dimensions shown in
Figure 3.2.1.
How do these five dimensions relate to the key components of mathemati-
cal literacy set out in the context-focused mathematics framework of Fig-
ure 3.1.1? TRU dimensions 1, 2, and 5 – content, cognitive demand, and
1 Content The extent to which classroom activities provide

opportunities for students to become knowledgeable,
flexible, and resourceful mathematical thinkers
2 Cognitive The extent to which students have opportunities to grapple

demand with and make sense of important ideas and their use
through productive struggle
3 Equity The extent to which classroom activity structures invite

and support the active engagement of all students
4 Agency, The extent to which students have opportunities to feel

ownership, and that the insights they develop are theirs and their fellow
identity students’, rather than just the teacher’s. This contributes
to their development of willingness to engage and of
positive identities as thinkers and learners
5 Formative The extent to which classroom activities and assessment

assessment elicit student thinking and how subsequent teaching
responds to those ideas, building on productive
beginnings and addressing emerging misunderstandings
FIGURE 3.2.1 The five dimensions of powerful classrooms.

Source: Adapted from Schoenfeld et al. 2016 [3A].
formative assessment – are central to both knowing how to use mathemati-

cal, data, and technology skills and knowing about modelling, data, and
technology – two of the five components of the framework.
TRU dimension 1 – content

Mathematical literacy draws on robust and flexible knowledge of mathemat-
ical content, built in classrooms that strongly exhibit these three TRU dimen-
sions. As noted in earlier chapters, sufficient ‘know how’ for mathematical
literacy means rather different things in different life-related contexts, and for
people at different levels of mathematical sophistication, but everyone ben-
efits from stronger understanding of mathematical and data content that they
have learned. Similarly, everyone benefits from having a good understanding
of how to use technology in solving mathematical problems. In Sections 3.3
and 3.4, we highlight some of the topics that are especially important for the
know how component of mathematical literacy. In Section 3.7, we address a
key aspect of the know about component (knowing about modelling), with
many examples in the later chapters.
Developing the know how to use mathematical modelling and the know
about to appreciate its use in society needs specific teaching. To build deeper
understanding of mathematics, the situations addressed need to include both
models that describe standard applications of mathematics and a variety of
non-routine real-world situations suitable for active modelling by the students,
some of which were described in Chapter 2. Life-related situations are often
messy; sorting out the variables that are important and identifying likely math-
ematical relationships between them is at the heart of mathematical literacy.
TRU dimension 2 – cognitive demand

Interesting life-related situations are often ‘open tasks’ where students can be
encouraged to take further the depth and extent of their analysis – both the
mathematical techniques used and the variables and data included. Taking
the context seriously creates a domain where cognitive demand at a level which
is conducive to productive struggle can arise naturally, as in many of the task
examples in Chapter 2. The need for critical thinking about methods and
results as part of mathematical literacy also raises the level of cognitive demand.
Of course, in classrooms, cognitive demand needs to be adjusted to a
workable level. Because of the strategic demands of model formulation, the
mathematical concepts and techniques that students can effectively deploy
in active modelling will be limited to those which they have thoroughly
absorbed, and connected to other aspects of mathematics and a range of
applications – usually topics that were first taught a few years earlier. This ‘few
year gap’ between imitative and autonomous deployment of mathematical
skills needs to be recognised in selecting situations to analyse; complex non-

routine tasks requiring recently taught mathematics will prove too difficult
for most students. Again, the core of mathematical literacy for most people is
Steen’s (2002) “sophisticated use of elementary mathematics”.
TRU dimension 5 – formative assessment for learning

This dimension emphasises that powerful instruction ‘meets students where they
are’ and gives them opportunities to deepen their understandings and, through
discussion in groups and with the class, ‘debug’ their errors and misconcep-
tions. With or without formal assessment, this dimension points to the need for
teachers to take time to examine how their students think about mathematics
in real-world contexts.
TRU dimensions 3 and 4 – equity, agency, ownership, and identity

These TRU dimensions are especially relevant to developing the context-
focused mathematical framework’s broader components of productive disposi-
tion and critical thinking. Classrooms in which a small number of students
get most of the ‘air time’ are not equitable, no matter how rich the content;
all students need to be involved in meaningful ways. Here open tasks can of-
fer different ‘entry levels’ and ‘end points’ to a problem. Students can begin
work using mathematical constructs with which they are comfortable and
then extend or refine solutions through group and individual work. Note of
caution: To foster a student’s agency, ownership, and identity as someone
who can tackle problems successfully, pointing out potential use of more
advanced mathematics in problem contexts should be retrospective, not un-
dermining the value of solutions that used simpler mathematics.
The essential role of the life-related context in mathematical literacy pro-
vides opportunities for students to bring their own specialised out-of-school
knowledge from hobbies and passions into a solution. Drawing on students’
special interests can help them become more engaged with mathematics be-
cause they see its relevance to topics which relate to who they are or are likely
to become. Even those who are usually unable to contribute much to conver-
sations in mathematics can make contributions, build on others’ ideas, and
have others build on theirs. This experience helps to develop a productive
disposition. Making comparisons between different people’s approaches and
evaluating mathematical insights against practical knowledge of the context
encourages critical thinking.
An approach where mathematical literacy problems put the student in the
role of a ‘consultant’, investigating the situation for a specific ‘client-audience’,
is powerful in developing agency and a productive disposition. It has the ad-
ditional value of requiring explanations of the analysis and of encouraging
FIGURE 3.2.2 Reducing road accidents – exploring the data.

Source: Bowland Maths [3B].
more thorough investigation. A consultant knows that a client needs a full

analysis and explanation. Reducing Road Accidents from Bowland Maths
[3B] is an example of a lesson that exploits this: students are tasked with pre-
paring a road safety plan for a small town. They are given a budget, a price
list of various schemes such as speed humps, pedestrian crossings, cycle lanes,
and education campaigns. They explore an interactive map-based database
of past accidents (Figure 3.2.2) to collect evidence. The series of lessons cul-
minates with students presenting their proposals to the class.
Some research (see e.g. Phillips et al., 1988) has shown that this kind of
role shifting is an effective way to get students to assume more autonomy and
responsibility for their work. Often it may simply be asking students to adopt
roles that are normally their teacher’s, such as explaining reasoning or asking
questions – questions that demand reasoned explanations, not just answers.
Changing the classroom contract

Taking the context seriously, our central component of mathematical literacy
presents some challenges for teaching and requires approaches that are new
in some classrooms. Often this implies a change in the ‘classroom contract’
(Brousseau, 1997). This refers to the usually implicit understanding between
teacher and students as to what each will do, and the roles each will play.
For example, in the traditional ‘3X’ or ‘pitcher’ model of ‘instruction’, the
teacher’s job is seen as filling the student up with knowledge – the teacher
eXplains a new procedure, works an eXample on the board, and gives the
students a set of eXercises involving closely similar tasks. The students expect
to watch the explanation, to follow the worked example, and then imitate the
procedure in doing the exercises. However valid the content is, this approach
is not strong on dimensions 2, 3, and 4 of TRU! Teaching for mathematical
literacy requires a ‘classroom contract’ where students have more autonomy
and responsibility for their learning and are expected to think critically about
what they do – a shifting of roles. Changing the classroom contract can be
hard – proceed in small steps, persist with new demands, and explain what
is different and why.
Reliable mathematics
Another aspect of the changed classroom contract involves reliability and
checking. The analysis of real-world situations often involves longer chains
of reasoning than most of the highly focused tasks that students tackle. If
any link in the chain is broken, through a procedural error, for example, the
subsequent reasoning and the conclusions are probably incorrect. In a set of
short exercises, 80% correct may be seen as “not bad”, but after four suc-
cessive steps with a random 20% error rate, you are more likely to be wrong
than right. More practice on short exercises is not the answer; instead, teach-
ers can raise students’ awareness of ‘checking tactics’. In increasing order of
the work likely to be involved, they are as follows:
• Sense making: Ask if the result makes sense in the context. If the result
(whether final or intermediate) seems surprising, then it is either very in-
teresting or (alas, more often) there was a mistake somewhere. In either
case, check it.
• Comparing different representations of the results: This can be a big help in
sense-making – graphs of various kinds may bring out unexpected features.
• Parallel calculation: Here, several students calculate separately, the results
are compared and, if there are discrepancies, the ‘bug’ is hunted down.
• Independent routes: Seeking to ‘do it a different way’ is a very power-
ful checking method. Finding a different way almost always gives further
insight into the context. Richard Feynman made this point to one of us:
“If you understand a result one way and it seems interesting, it’s worth
pursuing. If you understand it two ways, it may well be right. If you under-
stand it three ways, it almost certainly is”. – Richard Feynman, as related
by Hugh Burkhardt.
Of course, these checking methods only make the path from assumptions
to consequences more reliable; if the assumptions are wrong, at least you
know where to seek improvement.
A product
Creating a finished product from a substantial investigation into a real-world
situation is very valuable. It creates a focus for multi-step investigations and
is a concrete embodiment of a significant achievement by the student(s), re-
inforcing agency. The finished product may be a report, a presentation, or
an artefact. Displaying work facilitates cognitive and meta-cognitive com-
parisons with the work of others, which is itself valuable for learning. One
reasonably time-efficient way to organise this is a ‘gallery walk’, where stu-
dents view each other’s displayed work offering peer-review and construc-
tive feedback focused on specific features to help each revise their product,
perhaps leaving comments with sticky notes. The culmination of the Reduc-
ing Road Accidents lessons (Figure 3.2.2) was group presentations of their
accident prevention plans to the class, who represented the town council. The
presentations are supported by a poster or slideshow (examples are in the as-
sessment guide [3B]). Having a product is a pervasive and powerful feature
of Japanese mathematics classrooms, for example, which emphasize the ef-
fectiveness for learning of doing fewer activities in greater depth.
3.3 Orienting mathematical concepts

and skills for mathematical literacy
This section outlines some concrete suggestions for how teaching ‘normal’
mathematics content can be oriented to the teaching of mathematical literacy.
As we worked with the many mathematical literacy tasks that we considered
in preparing this book, we created the list below to summarise the con-
cepts, skills, and understandings that were most prominent. With just a little
change, these topics can be reoriented to better foster mathematical literacy.
Of course, this ‘little adaptation’ of content is not intended to be taught in an
abstract way divorced from the real world.
A mathematical literacy approach to number will emphasise the following:
• Using mathematical tools, especially calculators (on any device) and

spreadsheets.
• Quick estimation of the expected results of calculation.
• Calculating reliably (by whatever method) and checking results are sensible.
• Very large and very small numbers – their size and how to write them.
• Understanding and calculating with positive and negative powers of 10.
A mathematical literacy approach to measurement will emphasise the

following:
• The size of units in relation to real-world phenomena. (How fast does

1 metre per second feel? How many Joules, or kilowatt-hours, to boil a
litre of water?)
• Estimation of quantities, and knowing the typical measurements of some

common objects to use as benchmarks.
• Units for very large and very small quantities (e.g. terabytes of data, petajoules
of energy, nanometres) and their relationship to powers of 10 and 1,000.
• Reporting results of calculation with sensible precision.
A mathematical literacy approach to proportional reasoning will empha-

sise the following:
• All aspects of proportional reasoning and linking them (percents, ratios,

rates, etc.).
• Dealing with rates in less usual units (e.g. number of deaths per 100 000).
• Combining several rates (e.g. finding number of drops per minute to set
an intravenous drip using volume and time required and number of drops
per litre).
A mathematical literacy approach to statistics will emphasise the following:
• Four stages: reading the data, reading between the data, reading beyond
the data, and reading behind the data (Shaughnessy, 2007).
• Using technology (especially spreadsheet-like tools) to organise data sets,
create data displays, and do calculations.
• Increasing the range of data displays that students encounter in line with
changing practice (e.g. animated graphs).
A mathematical literacy approach to probability will emphasise the

following:
• Including discussion of risk to supplement work on chance.

• Linking mathematical with common language and practices (e.g. betting
odds).
• Visualising very small probabilities.
A mathematical literacy approach to geometry, spatial reasoning, and

location will emphasise the following:
• Identifying geometry in the world around and appreciating that geometric

features can make real objects work (e.g. a scissors lift keeps sections par-
allel, angles equal, and movement perpendicular to the ground).
• Using paper and digital maps and navigation tools.
• Visualising, interpreting, and making drawings and objects in three
dimensions.
• Specific links to hobbies or vocational interests of students.
A mathematical literacy approach to algebra will emphasise the following:
• Reading and using formulas.

• Ideas of independent and dependent variables.
• Behaviour of basic types of functions (linear, exponential, inverse proportion).
• That linear functions describe constant rate of change and exponential
functions describe a constant percentage rate of change (constant addition
or subtraction versus constant multiplication or division at each step).
Underlying these recommendations is an expectation that students learn

to use digital tools appropriately (see also Chapter 10) and that many of the
learning activities involve authentic contexts.
3.4 Tweaking mathematics lessons towards

mathematical literacy
In this section we look at three examples of how mathematics teaching prac-
tices can move more towards teaching for mathematical literacy. The first
is the adaptation of word problems, so that students can experience the
framework element of ‘taking the context seriously’. Next, we look at how
estimation (highlighted as an important emphasis under both number and
measurement in Section 3.3) can be included in almost any mathematics
lesson. The third example is about data displays, illustrating how curriculum
needs to adapt to changing mathematical practices in society.
Word problems to mathematical literacy problems

What do we mean by enriching tasks to make the link to mathematical lit-
eracy stronger? Here is a typical word problem on exponential decay.
The value of a car decreases by 15% a year. Anay buys a car for $10,000.
How much will it be worth after 6 years?
The solution expected is to identify and use the formula, FV = PV(1 − d)n,
where FV = future value, PV = present value, d = rate of decay per period,
and n = number of periods. In terms of our modelling process (Figure 1.3.1),
the model has already been formulated (exponential decay at a rate of 15%
a year) and all that is left to the student is the ‘solve’ stage – with minimal
‘interpretation’ and then ‘reporting’ the answer of $3771.495 to a specified
accuracy and remembering to include the dollar sign. Students encounter
many ‘word problems’ of this type where the task is essentially to spot the
right formula and calculate the answer. They are intended to strengthen skills
in identifying the formula, selecting the right data, and calculating correctly.
They do add to students’ appreciation of how formulas are useful in the real
world, in this case flagging that exponential decay is important for predict-
ing depreciation of assets. Stillman (1998) and also Brown (2019) label such
a real-world context as a ‘border’ context, which can almost be ignored by
the solver except to choose the formula giving loss of value rather than gain.
How could we enrich this task to better foster goals of mathematical liter-
acy? One change is to ask students to investigate the assumption of exponen-
tial decay (the missing formulation and evaluation steps). Another is to build
in some practical decision-making (bringing in interpretation and reporting).
The following version does both. In Stillman’s terms, this is a ‘tapestry’ con-
text, where the real-world task and the mathematics are woven together.
Anay buys a car for $10,000. Her friend Sam says a car loses 15% of its
value every year.
a Find a way to check if this model is realistic.

b According to the model, how much is the car worth after 1 year?
c Calculate the modelled value of the car over time – you might use algebra
or a spreadsheet.
d After owning the car for 4 years, it breaks down. Anay finds out that she
will need to replace the clutch to be able to drive the car again. How could
she find out if that is a good investment?
The curriculum probably requires students to practise the algebraic solu-

tion using the exponential function, but the second version permits the stu-
dent to use alternative methods (e.g. spreadsheet) and provides some support
in part (b). The second version also requires students to search for their own
data on used car prices and the cost to replace the clutch. All of this takes
longer than the first version (although data could be supplied when time is
short), but including a certain number of ‘tapestry’ problems amongst with
‘border’ problems strengthens a focus on mathematical literacy.
The original task has probably been formulated in that way to mimic the
assessment requirements: its focus on a single technical skill (solve using the
right formula); without access to everyday computing tools; a self-contained
question for an individual with no opportunity for research or discussion;
quick to complete – all leading to a single, easily scored right-or-wrong an-
swer. Importantly, there is certainly no need for teaching to be hamstrung by
those requirements by using the same style of task exclusively.
Estimation and appropriate accuracy in every lesson

This section examines estimation, to illustrate how the somewhat differ-
ent requirements of mathematical literacy can fit into and enhance the nor-
mal curriculum. Accuracy and precision are seen by many as a hallmark
of mathematics. Who else but mathematicians want to calculate millions of

digits of pi? But for mathematical literacy, quick estimation is often more
important. Estimation and selecting appropriate accuracy to report results
can play a role in many mathematics lessons.
Estimation is a multifaceted skill, relying on teaching in both number and
measurement. It is essential in all applications of any area of mathematics,
from quickly working out how much money to hand over to pay in a shop
and to check your change (estimation as rough calculation), to checking nu-
merical work (is my answer to these precise calculations reasonable), to buy-
ing enough lace to trim a dress (estimating a quantity), to providing a quote
to renovate a bathroom (combining many reasonable assumptions). When a
precise answer has been calculated (e.g. by hand or using a calculator), deci-
sions also need to be made about the sensible accuracy with which to report
the result. Is $3771.495 a reasonable answer for the value of Anay’s car after
6 years? What is sensible in that context?
Fortunately, nearly every lesson brings up some opportunities to estimate.
Quick checking of calculations is always important – in class as in real life. This
involves number-sense (is it reasonable that 5/13 is 3.6 – of course not, I ac-
cidentally divided 13 by 5 instead) and can be a creative activity, because there
are no prescribed ways to estimate. Instead, students can bring whatever they
know to estimate answers. They might check against physical experience: is it
reasonable that the third side of a triangle with two sides equal to 1 metre is
4 metres? Of course not, it is always shorter to walk directly from A to C rather
than going via B, so I expect an answer between 1 and 2. Of course, experience
is not always a good guide. For example, an internet app tells me the weight of
1 cubic metre of polystyrene is over 1 tonne. But polystyrene is so light! It does
not seem sensible: can this possibly be right? A mathematical literate person
will double or even triple check. Answer: yes, it is probably correct, but 1 cubic
metre of the polystyrene foam commonly used in packaging weighs only 50 kg.
Even that is too much for me to lift! A cubic metre is indeed a very large volume.
The two short formative assessment ‘smart::test’ [3C] items in Figure 3.4.1
involve estimation of length. Real-world estimation involves many skills,
each targeted by different items in the length tests and for which teachers
receive detailed reports (Stacey et al., 2009; Stacey et al., 2018). The top item
in Figure 3.4.1 tests knowledge of how long a millimetre is: students might
answer using ‘benchmark’ knowledge that a finger is about 1 cm wide. The
lower item is more complex, involving visualisation, estimates of length in
comparison to the ruler shown, and appreciation (from the multiple-choice
answers supplied) that rough calculations will give an adequate answer. I
counted about 20 horizontal segments, with estimated length of 50 cm (all
these lengths equal from the geometry), and four sloping segments of about
70 cm, leading to answer D. Items in a similar style in other smart::tests re-
port on estimation of numerical quantity, area, mass, and volume.
The length of a grain of rice is

A. less than 2 mm
B. between 2 mm and 10 mm
(correct)
C. between 10 mm and 20 mm
D. between 20 mm and 30 mm
E. more than 30 mm
This picture shows a clothes airer

made of two different thicknesses
of plastic-coated wire. There is a
30 cm ruler on the floor.
The total length of the thinner
wire is
A. less than 2 metres
B. between 2m and 5 m
C. between 5 m and 10 m
D. between 10 m and 15 m
(correct)
E. more than 15 m.
(Modified from original to
account for the black & white
image)
FIGURE 3.4.1 Two SMART::test items on estimation of length.

Source: Smart::tests Estimation of Number (Quiz A) Items 4151 (top) and 4282 (bottom) [3C].
Classroom discussion of even short items like these can illustrate many
points about real-world estimation:
• That the desirable precision in an estimate depends on the real-world

purpose of the answer (in these items – detached from a real-world
purpose – precision is only indicated by the range of multiple-choice
answers provided as in Space Flight task in Figure 2.1.1).
• That speed is important: estimation is used when detailed counts,

measurements, and calculations are too slow, not required, or perhaps
impossible.
• That the base knowledge for estimation is an appreciation of the size of
measurement units and the relative size of numbers, with enough fluency
with number facts (including power of ten) and the effect of operations to
estimate the results of calculations.
• That other cognitive skills related to the context are needed to contribute
to estimation, for example, three-dimensional visualisation for the clothes
airer item.
• That a mathematically literate person knows a range of benchmarks for
common units of measurement (for example, they might know that the
height of a tall man is about 2 metres, that 1 litre of water weighs a kilo-
gram, that a finger is about 1 cm wide, and that a can of drink holds about
400 mL).
Students can learn strategies for estimation, many of which involve pro-
portional reasoning (especially scaling known benchmarks up or down).
Statistics – expanding the range of data displays

Whilst the school curriculum in many countries may still be focused on picto-
grams, bar charts, and pie charts, new software and new capabilities of digi-
tal publishing have prompted the growth of new and more informative data
displays. Even if a mathematics curriculum only requires the students know
how to create traditional data displays, students dealing with data from other
areas of interest will need to interpret new graph types.
Figure 3.4.2 presents one example of the new types of data displays that
are now being used. It is adapted from the outstanding Gapminder website
[3D]. This site presents a huge and constantly updated database of statistics
about people and countries around the world. The main messages are that
facts (data) are powerful, that things have been getting better over time for
most people, and that people around the world who have similar incomes
have reasonably similar lifestyles no matter where they live (Rosling et al.
2016). In addition to promoting these powerful messages, the Gapminder
website has pioneered and popularised new ways of displaying data. Tradi-
tional data displays taught at school (e.g. pie chart or line graph) show the
relationship between only two variables, but the outcomes of concern to the
Gapminder website are often linked to many variables, necessitating a more
complex display. Figure 3.4.2 shows how a ‘bubble chart’ can combine GDP
per capita (horizontal axis), life expectancy (vertical axis), population (size
of bubble), and country (label). This figure has been drastically simplified
here to meet the limitations of the printed page – the live online version [3D]
FIGURE 3.4.2 Life expectancy versus per capita income in 2022 compared to 1822.
Source: Free material from Gapminder [3D].
includes far more countries, adds colour to show the world region, and
motion to show the passage of time, ultimately combining six variables on one
chart, and allows the data for any country to be viewed by pointing to it. This
data display powerfully illustrates that life expectancy is strongly linked to
average income and that all countries where data is available have improved
on both variables over time. Compare the 2022 with the 1822 data of the
inset on the figure. Because the ‘bubbles’ represent country population size,
the display also shows that the improvement has been for people, not just
for countries. Mathematically literate citizens can easily investigate other re-
lationships of interest by selecting their own variables and data sets to better
understand our changing world. Further classroom activities for exploring
data are offered in Chapters 4–10 of this book.
3.5 Deep engagement with the context

through interdisciplinary work
As we have noted many times above, a distinguishing feature of teaching
mathematical literacy must be to help students to intertwine their under-
standing of the context and the mathematics. Working with teachers of other
subjects in a secondary school or using cross-curriculum investigations in a
primary school can provide excellent opportunities to deeply intertwine an
authentic context with mathematical thinking. Done well, this strengthens

students’ appreciation of mathematics and its applications (‘know about’
as well as ‘know how’), improves some switched-off students’ disposition
towards mathematics, and also gives serious attention to the mathematical
concepts and skills that are needed in real contexts using the language and
practices of the area of application.
To do this well, we need to base lessons and projects on very well-selected
topics. Much mathematical work outside of the mathematics classroom
is very routine, such as adding numbers, and does not involve significant
mathematical thinking. Minimising mathematics removes obstacles for the
teacher trying to teach difficult concepts in another field, but it does not assist
with developing mathematical literacy. Fortunately, there are many good
resources that suggest interdisciplinary lessons with substantial content on
both sides (see e.g. Goos et al., 2019); below we describe two such lessons.
Planning for interdisciplinary teaching requires some degree of cross-
subject planning. As noted above, it is nearly always important that the math-
ematical concepts and skills involved are already familiar to the students,
because learning about new content will take considerable ‘mental space’,
distracting from focus on the context. Working with teachers of other sub-
ject areas has the side benefit of enriching mathematics teachers’ own under-
standing of the different ways in which mathematics is used, and the varying
mathematical practices and language that their own students have to learn
for different school subjects, disciplines, trades, and jobs.
In summary, even if they are not directly involved in teaching the interdis-
ciplinary lessons, mathematics teachers can encourage and support teachers
of other subjects to include significant mathematical content, ensure the rel-
evant content is already well understood, and look for interdisciplinary op-
portunities that will help their students become more mathematically literate.
Isometric drawing of licorice allsorts

This lesson resource comes from the Numeracy across the Curriculum pro-
ject of the Department of Education in Victoria, Australia [3E]. The website
provides a range of sample lessons looking at the mathematical literacy
demands of most school subjects for students aged 11–16. Each lesson is
prefaced by a short survey of the main mathematical demands of the subject
at the relevant stage of the curriculum, which is useful for both mathematics
teachers and the subject teacher.
This lesson on isometric drawing is written for teachers of the subject
Visual Communication Design for students 12–14 years of age. Isometric
drawing is a straightforward drawing technique that is useful to introduce
students to pictorial drawing prior to teaching more difficult technical draw-
ing techniques. It involves one projection of three-dimensional space onto
two-dimensional space – one of the big challenges for artists and map makers
throughout the ages. This lesson is not about the rich mathematical history of
projections, but directly teaches students the rules for isometric drawing, the
skills involved, and how it can have visual impact. Full details of this lesson
are available online [3E].
A licorice allsort is a delicious sweet, made of two square layers of licorice
sandwiched between three brightly coloured sugar paste layers (see Figure
3.5.1). In the lesson, students look at objects drawn with several different
projections and compare the visual impact. Licorice allsorts are good to draw
because they have a simple and important three-dimensional shape, they are
colourful, and each student can have their own allsort to measure and to
view from all angles. Students enlarge the real measurements in the ratio
of 5:1, so the drawing fits well on A4 paper. They centre the drawing on
the paper using their own measurements and learn how to show the layers
of the licorice allsort on the three isometric axes. Students experiment with
colours and texture before choosing how to render the layers with coloured
markers for their final artwork. They use fine liners to outline the drawing
and the border. The finished pictures can be assembled to make an attractive
classroom display.
Success in this activity depends on knowledge of multiple mathematical
concepts and skills, listed in Figure 3.5.1. In turn, students’ understanding of
these aspects of mathematics is strengthened by seeing them in action in this
important context. Furthermore, a practical demonstration of the usefulness
of mathematics in activities students enjoy can help build positive disposi-
tions to learning mathematics.
Required mathematical knowledge:

• Knowledge of 2D shapes and 3D objects, including their names
• Identification of parallel and perpendicular lines in two and three dimensions
• Measuring lengths of lines and size of angles (30 degrees and 60 degrees)
• Enlarging shapes in the ratio 5:1
• Using measurements to locate the exact centre of the A4 paper
FIGURE 3.5.1 Stages of an isometric drawing and mathematical knowledge required.

Source: Numeracy across the Curriculum – Victoria Department of Education [3E].
Measuring biodiversity with probability

Biodiversity is important for ecosystem health and productivity. High bio-
diversity makes ecosystems more resilient to stresses, because having more
species means that there is a better chance that some organisms will have
traits that are necessary to survive or adapt. To make decisions such as de-
ciding which of several locations is the best place for a marine conservation
park or a road, or whether a conservation programme has been effective,
ecologists need a practical and reliable measure of the biodiversity of an eco-
system. The lessons highlighted in this section focus on Simpsons’ Diversity
Index. This is used as a measure of diversity in many biological, economic,
geographical, and demographic contexts, and often used in biological sub-
jects in Years 9–12; see, for example, the UK A-level resource website by the
Royal Geographic Society [3F]. Because the mathematics behind the index is
accessible to students who have studied some probability, it makes an excel-
lent interdisciplinary module, as explained by Duncan et al. (2014). Simpson
introduced his index in 1949 and (confusingly) there are several variations
in common use, some named after other people. Several online Simpson’s
Diversity Index calculators are available [3G].
Designing a measure for biodiversity is at heart a challenge to design a
mathematical model that captures the key factors, uses obtainable data, and
combines the information to produce a number that is meaningful. (This is
discussed in Chapter 12.) The module by Duncan et al. (2014) begins by
discussing the need to quantify biodiversity, and then illustrates the two key
concepts: species richness and species evenness. Species richness is the num-
ber of unique species in the area being studied – generally, the greater the
number of species, the more biodiverse the area. Species evenness gives a
measure of how much variation there is between the number of individuals
in each species – generally there is more evenness in more biodiverse areas.
Normally, the diversity index is calculated for a group of similar organisms
rather than the whole ecosystem, so evenness is calculated from the number
of trees of different species (for example), rather than comparing the very
disparate numbers of trees at the location with the numbers of ants. To cal-
culate Simpson’s index, the species of interest in the area are ascertained and
the number of individuals of each species is established (using sampling for
large sites). Simpson’s Diversity Index is defined as the probability that two
organisms selected at random from the area are of different species.
Understanding how the formula is derived is within the mathematical ca-
pability of these senior students. Students (probably working in groups) can
collect their own data, say to compare two sites, and calculate the indices.
Alternatively, they can use published data sets. Students then need to ex-
plore the meaning of the index numbers that they have calculated, and how
they reflect species richness and evenness. The index is one example of a
widespread phenomenon of constructing mathematical functions to act as

measures of abstract constructs. Citizens come across these daily. Chapter 12
looks at the mathematical literacy demands this imposes.
This lesson illustrates a number of features about mathematical literacy.
Understanding many aspects of the context is essential to understand the goal
of the index, why it is needed, and why it works. A teacher of ecology or ge-
ography or biology is needed to assist students to understand what variables
related to ecosystems might be significant and should be included somehow
in the index, how data sets should be assembled, and what data can realisti-
cally be collected and how. At first glance, the only mathematics required is
to be able to substitute into the formula (or use an online diversity calcula-
tor!), but understanding how the formula is derived helps with interpreting
the numerical results, as well as demonstrating an important and unusual
application of probability theory with a genuine link to students’ other stud-
ies. Students can explore the results produced by the formula by seeing how
extreme situations of richness or evenness influence the index, thereby getting
a sense for the strengths and limitations of using such an index (a mathemati-
cal model!) for making decisions. Other indices of diversity can be compared.
Critical thinking is employed when students provide evidence-based argu-
ments based on the index to support their decisions and to evaluate the ar-
guments of others – and critique the strengths and weaknesses of the index.
3.6 Teaching for critical thinking

Previous sections in this chapter are mostly concerned with what is some-
times called ‘functional numeracy’: how mathematical concepts and skills
can be applied in the solution of everyday or workplace problems. However,
meeting the intention that mathematical literacy gives people power in their
lives requires more than traditional mathematical skills. People also need the
capability to think critically about problems and their possible solutions, to
carefully evaluate the strength of the evidence and examine alternative pos-
sibilities, and to consider the implications of findings from a personal, work-
place, or social and community point of view. Mathematical models and the
outcomes of using them need to be critiqued considering the variables in-
cluded and omitted, the relationships proposed between them, the quality
of evidence for the assumptions, and the interpretation of results. Critiquing
Simpson’s Diversity Index as a measure of biodiversity is one example. Stu-
dents also need the productive disposition to do so.
How do we help students develop a critical orientation? This has been
investigated by many scholars, including Andersson and Barwell (2021). As
for all higher order thinking skills, students acquire experience from active
learning pedagogies, where they are challenged to form and explain argu-
ments and justify their conclusions while working on problems with some
degree of autonomy. Working with others when analysing information and

forming arguments and evaluating conclusions means that more ideas can
be considered and ideas can be tested as the work proceeds. Having stu-
dents present their work to others gives opportunities to practise evaluating
arguments and assessing the strength of the evidence presented. For math-
ematical literacy, this will often include thinking carefully about how well
the situation has been modelled with mathematics – how appropriately the
assumptions and relationships that have been built into the model capture
what is important.
Critical thinking is also likely to be prominent in drawing conclusions from
data, especially when they go beyond the actual data that has been presented.
Chapter 5 on climate change provides instances of both of these aspects.
Because of scientific critique, scientists continually refine their climate models
so that they can incorporate more and better input information to improve
the models. Critical thinking is also important for governments, business,
and citizens to understand how to interpret and act on the outcomes. Class
discussion is needed to help students learn from their experiences, with the
teacher taking a leading role in highlighting key moments and ensuring that
the focus remains on asking and answering sensible and important questions
about the context. The later chapters of this book provide many examples
where the deeper analysis of context enables informed critical thinking.
Controversial and sensitive issues

A consequence of adopting a critical pedagogy is that teachers will need to
manage some potentially controversial or sensitive issues in their classrooms.
If students are really to consider the social or personal or environmental
consequences of the data and models that they study and are encouraged
to see their knowledge as giving them power in their lives, then feelings can
run high. Mathematics teachers who are moving into this terrain often find
it helpful to talk to teachers of subjects in the humanities or social sciences,
who have long faced this kind of challenge.
Thinking critically about mathematical results and their implications will
raise issues of values and ethics, and political questions about how society
works. If mathematical literacy does indeed support decision-making, teach-
ers can expect that controversial conclusions will sometimes arise about the
actions that should follow. Within one class, students who have different
experiences and background and different values may well draw different im-
plications from mathematical results. But these conversations are important
for helping students become informed and active citizens, surely an impor-
tant goal of education. The most famous example of a student acting on what
they learned at school is Greta Thunberg who began a worldwide movement
of ‘School Strike for Climate’ protests.
Some students may disagree strongly with the course of action others rec-
ommend, perhaps heatedly. In dealing with these disagreements, mathemati-
cal literacy is best developed by keeping the focus of class discussion on the
quality of argument presented: are the premises of the argument sound (in-
cluding the data), are the steps of the argument logically valid, how does the
conclusion of the argument lead to the recommendation for action, and what
might the consequences of such action really be.
The Gapminder website [3D], discussed in Section 3.4, is a powerful tool
for developing students’ ‘global awareness’ and their knowledge of basic in-
dicators of well-being. In our highly connected and quickly changing world,
this is important. But just behind the data lie many controversial issues: the
phenomenal success in moving people out of poverty in China balanced
against political freedom, lingering effects of colonisation, the economic im-
pact of female illiteracy, the extent of women’s rights, and the miserable ef-
fects of war are just a few. Data from any one country will raise different, but
still controversial, issues. Using real social data in the mathematics classroom
and expecting students to think critically about it inevitably means that some
of the big debates about society are encountered.
Some issues may be especially sensitive for some classes (if, for example,
a classmate has recently died) and are best avoided, whilst being potentially
beneficial for most classes. Teachers, knowing their students, make the judge-
ment on this, and perhaps seeking advice from student welfare or health
education teachers. Chapter 4 raises various issues of this kind – for exam-
ple, data shows that the death rate for people aged 15–25 is much higher for
males than for females. Discussions on money in Chapters 7 and 9 need to
be handled in a way that recognises that different students in a class will have
very different wealth, opportunities, and values.
3.7 Knowing about modelling and thinking

critically about models
This section draws together some ideas about mathematical models that have
underpinned most of the examples in these first three chapters. An important
part of the ‘know about’ component of mathematical literacy is to understand
the basic ideas behind mathematical modelling – the process of drawing on
an entirely abstract world of mathematical concepts, to make predictions and
explain real phenomena in the physical and social worlds. Helping students
develop this appreciation, so that they know what can be expected by using
a mathematical model, and what cannot, is part of teaching for mathematical
literacy. Appreciating what mathematical modelling can do depends on under-
standing its limitations and its strengths. Only a small proportion of people
will ever create any but the simplest of mathematical models, but we are all
users of them, so the need is more often to ‘know about’ than to ‘know how’.
A first point to teach is that people encounter many mathematical models

even in their everyday lives. Rules for cooking are frequently examples of
simple mathematical models. My rule for the absorption method for cook-
ing rice is to add 1.5 cups of water for every cup of rice: easy to remember
(a great advantage) and works well, although some people recommend two
cups of water instead. Does it make sense to say one of these models for pre-
dicting well-cooked rice is right and the other is wrong? Not without looking
at the assumptions behind the recipe such as the type of rice and also thinking
about the criterion of success – what counts as well-cooked rice for you? Per-
haps they are both ‘good enough’, so a home cook can remember either one.
Students also need to know that many ‘apps’ that people use every day are
built on mathematical models (yes, apps to help with cooking rice are avail-
able), some extremely complex. When I use an online navigation app to go to
work, the system finds several possible routes, drawing on its extraordinarily
detailed map of my town. This map is really a digital model incorporating
the geography and the road system. It divides the route into segments, based
on road type or speed limit. Then it uses the speed limit and/or real-time data
about speed gathered from the phones of other users currently driving to make
a recommendation for my route. If I want to go at 6 a.m. on Friday morning
instead, the model will replace the real-time data from cars with accumulated
average speeds for that time. The model is generally surprisingly accurate, but
no one expects it to be perfect. It cannot predict that an accident that might
soon block an intersection. When evaluating a model for most situations, we
ask whether it is often very helpful, not if it is exactly correct.
Mathematical models are most often invisible to end users. For example,
social media platforms sell advertising, and can earn revenue when users
click on an advertisement. Platforms employ machine learning to identify
detailed profiles (mathematical models) of users who are likely to click on
certain types of advertisements. By targeting advertisements to those most
likely to click on them (and especially those who will ‘like’ them), the plat-
form can maximise clicks and hence revenue. For news and information,
there is research that indicates that negative reactions (outrage) yield more
clicks than positive reactions, encouraging the design of algorithms in the
routing models to offer more provocative links. There are claims that this
has led to widespread harm by polarising social and political attitudes and
spreading pornography. A mathematical literacy critique of a model like this
might consider its effectiveness in both identifying the most valuable users
(the adequacy of the model) and a social critique of the consequences (in-
tended, unintended, certain, possible, good and bad). The uses and misuses
of mathematics in politics, media, and business are discussed in Chapter 7.
Critical thinking is part of mathematical literacy.
A mathematically literate person needs to appreciate how much our so-
ciety depends on mathematical models and to appreciate just a little of the
complexity behind some of them. To issue a forecast, a computer program

for weather prediction works with an enormously complex model – a set of
mathematical relationships between hundreds of variables, with millions of
input data from land, sea, and space. Our common experience is that the
weather predictions are sometimes ‘wrong’ (rain might not come when it
is predicted) but very ‘useful’ both for trivial decisions (will I take my coat)
and for immensely important decisions (should this town be evacuated to-
night because people are in the predicted path of a typhoon). Many people
are continuously employed to improve the models to get better forecasts. A
good model gives a useful prediction most of the time – perfection is not an
option. Another type of mathematical model is to make a measure of an ab-
stract construct for entirely practical purposes, as Simpson’s Diversity Index
in Section 3.5 does. Gross Domestic Product is another example. Such meas-
ures can be used to compare, to track over time, and as a basis for decision-
making. They are the topic of Chapter 12.
In summary, we recall the famous saying that “all models are wrong, but
some are useful” (Box, 1979); so useful indeed, we now cannot live without
them.
3.8 How do we move forward?

In this chapter we have set out implications for teaching that addresses the
components of mathematical literacy that we outlined in Chapter 1. The
focus on taking the context seriously has been present throughout, with an
emphasis on taking a ‘critical thinking’ approach to data and in using and
understanding mathematical modelling and models, whose range, ubiquity,
and dependence on technologies is outlined in Section 3.7. This is comple-
mented by the suggestions in Section 3.3 and 3.4 for developing a range of
mathematical skills in specific areas like estimation that are especially useful
for real-world applications. Sections 3.1 and 3.2 outlined two ‘fives’: the
five key components of the context-focused mathematical framework that
learning activities must address to develop mathematically literate thinkers
within the broader pedagogical context of TRU’s Teaching for Robust Un-
derstanding, with its five dimensions.
For many teachers of mathematics, this agenda represents a qualitative
enlargement of the range of learning goals to which they are accustomed.
They can form part of ongoing professional development through local dis-
cussion, formal professional learning sessions, or community models such as
lesson study. There is online support for such professional development, such
as the Bowland Maths professional development modules [3H]. These offer
a sequence of structured activities for a group of teachers to work together
on specific pedagogical challenges. The design is a three session ‘sandwich
model’ of classroom-activity-based professional development, designed to
maximise ‘carry over’ from the sessions to the classroom – worlds that can
sometimes seem far apart. Teachers first work together to plan a lesson sup-
ported by supplied materials, then teach and observe the outcomes at school,
and then come together again to reflect on the activity and what they have
learned. The module handbooks describe the activity sequence in detail and
address topics such as tackling unstructured problems, assessing modelling
processes, and managing collaborative work.
Acknowledgements
Figure 3.4.2 is based on free material from www.gapminder.org available un-
der the Creative Commons Attribution 4.0 International licence. Figure 3.5.1
is adapted from material from Numeracy across the Curriculum © State of
Victoria (Department of Education and Training), also available under the
Creative Commons Attribution 4.0 International licence. The image from
Bowland Maths (Figure 3.2.2) appears courtesy of the Bowland Maths main-
tainers. Material from smart::tests in Figure 3.4.1 appears with permission
from Kaye Stacey.
References
Andersson, A., & Barwell, R. (2021). Applying critical mathematics education.
Brill.
Box, G. E. P. (1979). Robustness in the strategy of scientific model building. In R. L.
Launer & G. N. Wilkinson (Eds.), Robustness in statistics (pp. 201–236), Aca-
demic Press. https://doi.org/10.1016/B978-0-12-438150-6.50018-2.
Brousseau, G. (1997). Theory of didactical situations in mathematics: didactique des
mathematiques, 1970–1990. Kluwer.
Brown, J. P. (2019). Real-world task context: Meanings and roles. In G. Stillman &
J. Brown (Eds.), Lines of inquiry in mathematical modelling research in education.
ICME-13 Monographs. Springer. https://doi.org/10.1007/978-3-030-14931-4_4
Burkhardt, H. (2021). Modelling in school mathematics: Past achievements – current
challenges. In G. A. Stillman, W. Blum, & G. Kaiser (Eds.), Mathematical model-
ling and applications, ICTMA 19 (pp. 529–539). Springer.
Duncan, S. I., Lenhart, S., & Sturner, K. K. (2014). Measuring biodiversity with
probability. The Mathematics Teacher, 107(7), 547–552. https://doi.org/10.5951/
mathteacher.107.7.0547
Goos, M., Geiger, V., Dole, S., Forgasz, H., & Bennison, A. (2019). Numeracy across
the curriculum: Research-based strategies for enhancing teaching and learning (1st
ed.). Allen & Unwin.
Madison, B. L., & Steen, L. A. (Eds.), (2008). Calculation vs. context: Quantitative
literacy and its implications for teacher education. Mathematical Association of
America. https://www.maa.org/sites/default/files/pdf/QL/cvc/CalcVsContext.pdf
Phillips, R., Burkhardt, H., Fraser, R., Coupland, J., Pimm, D., & Ridgway, J. (1988).
Learning activities and classroom roles with and without the microcomputer.
Journal of Mathematical Behavior, 6, 305–338.
Rosling, H., Rosling, O., & Rönnlund, A. R (2016). Factfulness: Ten reasons we’re
wrong about the world – and why things are better than you think. Sceptre.
Shaughnessy, J. M. (2007). Research on statistical learning and reasoning. In F. K.
Lester (Ed.), Second handbook of research on mathematics teaching and learning
(pp. 957–1009). Information Age Publishing.
Stacey, K., Price, B., Steinle, V., Chick, H., & Gzodenko, E. (2009). SMART assess-
ment for learning (paper presentation). International Society for Design and Devel-
opment in Education Conference 2009, Cairns, Australia. Retrieved from https://
www.isdde.org/conferences/conference-cairns-2009/working-groups-and-themes/
curriculum-documents-assessment/
Stacey, K., Steinle, V., Price, B., & Gvozdenko, E. (2018) Specific mathematics assess-
ments that reveal thinking: An online tool to build teachers’ diagnostic competence
and support teaching. In T. Leuders, K. Philipp, & J. Leuders (Eds.), Mathematics
teacher education: Diagnostic competence of mathematics teachers (pp. 241–261).
Springer.
Steen, L. A. (Ed.). (2002). Mathematics and democracy: The case for quantitative
literacy. National Council on Education and the Disciplines.
Stillman, G. (1998). The emperor’s new clothes? Teaching and assessment of math-
ematical applications at the senior level. In P. Galbraith, W. Blum, G. Booker,
& D. Huntley (Eds.), Mathematical modelling: Teaching and assessment in a
technology-rich world (pp. 243–253). Horwood.
United Nations Sustainable Development Goals for 2030 (n.d.) Target 4.6. Retrieved
from: https://sdgs.un.org/goals/goal4

e.g. ltml.mathlit.org/3A
purposes.
[3A] The TRU Framework – Teaching for Robust Understanding Project

https://truframework.org/
[3B] Reducing Road Accidents – Bowland Maths
https://www.bowlandmaths.org.uk/projects/reducing_road_acci-
dents.html
[3C] Estimation of Length and Mass Items – SMART::tests
https://www.smartvic.com/
[3D] Understand a Changing World – Gapminder
[3E] Isometric Drawing – Numeracy across the Curriculum
https://www.education.vic.gov.au/school/teachers/teachingresources/
discipline/maths/Pages/numeracy-for-all-learners.aspx#link27
[3F] A Guide to Simpson’s Diversity Index – Royal Geographical Society

https://www.rgs.org/schools/resources-for-schools/a-student-guide-to-
the-a-level-independent-investigation-non-examined-assessment-nea
[3G] Simpson’s Diversity Index Calculator – Statology
https://www.statology.org/simpsons-diversity-index-calculator/
[3H] Professional Development Modules – Bowland Maths
https://www.bowlandmaths.org.uk/pd/
4
HOW RISKY IS LIFE?
Risk is an area of life that concerns many people – and rightly so. Numbers
are obviously involved, so it is an important aspect of mathematical literacy.
But risk is also an area where many people’s awareness and concern have
little connection to the level of risk actually involved in a situation or an
activity.
This sometimes has life-changing consequences. Parents are reluctant to let
their children ‘play out’ which undermines their physical and mental health
and development. Inadequate physical activity reduces general fitness and
increases the risk of obesity and diabetes – screen games are no substitute.
People have a fear of flying or other phobias, where the effects on their career
and lifestyle may be significant. Yet the data shows that the likelihood of harm
in such cases is extremely small. These are just two examples where increasing
mathematical and, in particular, data literacy can improve quality of life.
These are emotional subjects that make it hard for deliberative ‘slow
thinking’ to cut through instinctive reactions to news reports of tragic events,
and thus to improve decisions. The media know that people like stories –
but stories are not data from which you can make reliable judgements. Any
discussion of risk must take this into account.
The theme of this chapter is that people can make better-informed deci-
sions about what to do about potentially risky situations if they take the
trouble to make informed estimates of the likelihood of the risk. With a bit
of searching, adequate data is often available on the web. Learning to do this
kind of deliberative thinking, and being confident in doing it (the produc-
tive disposition component of the context-focused mathematics framework
of Chapter 1) is an aspect of mathematical literacy with a big potential payoff
in quality of life – one that many find interesting, and motivating.
DOI: 10.4324/9781003303503-5
How risky is life? 77
In Section 4.1 we set out a general framework for thinking about risk,
moving on in Section 4.2 to describe an approach to developing understand-
ing of risk in the classroom, based on well-engineered materials that are
freely available online. Section 4.3 looks at the challenges of finding reli-
able data and probing it for other undeclared influences. The COVID-19
pandemic made people more aware of risk – the chance of getting infected,
and of the consequences. Section 4.4 looks at this complex example in some
detail, pointing out general features and lessons learned that may help prepa-
ration for the next epidemic which, with unknown effects and mechanisms, is
expected in a decade or so. Section 4.5 looks at the issues in medical testing –
an example of conditional probability directly relevant to the mathematical
curriculum where there are surprising results. Again, effective lesson mate-
rials are available free online. The chapter includes a mix of ‘know how’
and ‘know about’ the analysis of risk, using a critical thinking approach.
It also provides many examples of using both the theory-driven and data-
driven a pproaches to u nderstand a situation. Risk is a topic, that interests
almost everyone, helping to build a productive disposition to investigation
in general.
4.1 Dimensions of risk

There are two dimensions of risk that need to be distinguished. Any potential
event has both a likelihood and a level of hazard:
• Likelihood: How likely is this to happen?

• Hazard: How serious is it, if it does happen?
These two aspects both need to be taken into account, and balanced, in
making decisions. In this chapter we use ‘likelihood’ as the general term,
reserving ‘probability’ as a measure of likelihood in its formal mathematical
sense – a number in the range 0 (never happens) to 1 (for sure).
Some events are high on likelihood but low on hazard. I often forget where
I left my keys –but I only waste a few minutes searching for them. I’m likely
to get a cold this winter, but it usually only means a few unpleasant days. So
I won’t change my lifestyle to reduce the likelihood.
Other events are very unlikely but disastrous if they do occur. The chance
of a large meteorite hitting the earth is extremely small, but the one that
plunged into the Gulf of Mexico 66 million years ago sent so much stuff
into the atmosphere that the whole ecology was changed for long enough to
kill off the non-flying dinosaurs (others became birds) and create ecological
‘space’ for mammals to increase and diversify – which might be seen as pro-
gress. Scientists now track objects in space with a view to diverting any that
pose a risk like that. On hazard at a personal level, each of us has in mind
78 How risky is life?
possible life-changing events that we regard as disastrous – someone close

to us dying is an obvious example; we take what steps we can to lower the
likelihood of that happening.
Most potential events are in between these extremes – with fairly low
likelihood and consequences that are unpleasant but not life-changing. Risk-
ing breaking your ankle on a school outdoor adventure trip, or a parachute
jump, or riding your bicycle to school every day to improve your fitness pre-
sent risks; conversely, not exercising regularly is also a major risk to health –
as is spending every evening playing video games. These choices all have risks
to health and well-being and there is data that enables you to estimate their
likelihood, at least for the average person.
These examples illustrate another distinction that is important in thinking
about risk: between ‘one-off events’ like the parachute jump and ongoing
activities like riding a bicycle to school.
Because quantifying hazard involves subjective judgements, people differ
in how serious they regard the hazard to be. In much of this chapter. we shall
‘fix’ the hazard variable by concentrating on the likelihood of various fatal
outcomes. Death also provides a background frisson in classroom discussions
that, though it needs sensitive handling, students seem to enjoy.
Measures of risk to life

R.A Howard (1980) introduced a useful unit for comparing risks to life, the
micromort: a one in a million chance of death – or probability 0.000001 = 10−6.
It is a useful unit because long decimals are hard to understand.
ACTIVITY
Discussing in small groups, conjecture how many micromorts (chances in a

million) is the average likelihood of dying in: a parachute jump; giving birth;
a car accident during one year.
Table 4.1.1 gives some examples for individual events, and for ongoing
activities like riding or driving in wealthy countries.
Base risk
Another important number in thinking about risk is the total probabil-
ity that an average person in a specified group will die of any cause in the
coming year. We call this ‘base risk’. The most important variable affect-
ing base risk is age – though many other variables, including health, gender,
TABLE 4.1.1 Likelihood of dying in various ways in wealthy countries
Hazard Chance of death
Single event
A skydive/parachute jump 8 micromorts
Climbing to the top of Mt Everest 40,000 micromorts
Giving birth 150 micromorts
Baby’s first day of life 430 micromorts
For ongoing activities
Being murdered 10 (UK) to 50 (US) micromorts per year
Car accident 25 micromorts per 10,000 km
Travelling by train 1 micromorts per 10,000 km
Motorbike accident 1000 micromorts per 10,000 km
All unnatural causes 230 (UK) to 460 (US) micromorts per year
Source: Data from Wikipedia and other sources.
socio-economic status, occupation, and country, will influence it, sometimes

substantially. Base risk as a function of age is a component of the ‘life tables’
published by national statistics authorities. (You might think ‘death tables’
is a better name!) Table 4.1.2 is a recent example for England and Wales,
also showing the effect of gender. The last column gives the correspond-
ing e xpected age at death – numbers that are perhaps surprising at first but
obvious on reflection.
TABLE 4.1.2 Probability of death in coming year, and expected age

at death
Current age Probability (micromorts) Expected age at death
Males Females Males Females
0 4,220 3,500 79 83

1 229 214 79 83
5 74 74 79 83
10 78 66 79 83
20 525 187 80 83
30 771 387 80 83
40 1610 924 80 84
50 3580 2,230 81 84
60 7910 5250 83 85
70 19 200 12 900 85 87
80 56 100 39 600 88 90
90 163 000 136 000 94 95
Source: Calculated from UK Office of National Statistics tables England and
Wales 2018–2020 [4A].
It may seem strange that the chance is not higher at the older ages – only
about 15% (150 000 micromorts) chance of dying within a year at age 90.
But we can estimate, say, the 5-year survival probability at age 90. The prob-
ability of surviving 5 successive years (1 − 0.15)5 = 0.44, which makes it
plausible that 95 is the expected age of death.
One of the interesting features of the COVID-19 pandemic was how closely
the chance of death from a COVID infection paralleled the base risk as a func-
tion of age – increasing by a factor of around 2.5 for every 10 years from age 30.
In considering how seriously to take a risk, your base risk is a useful
comparator. Not that we are recommending parachute jumping for 80-year-
olds – but, if you fancy it, sky-diving is a relatively minor risk.
The base risk of dying and the danger of various hazards provide the back-
ground to answering one of the questions with which we began this book:
What is the chance that a teenager (like my child) will die in a shooting at
school? How does it compare to other hazards in their life?
In 2022, 32 students were killed in 51 school shootings in the United States.

There are 115 000 US schools that educate 50 million students across the age
range 4–18. So the risk of dying in a school shooting in 2022 was less than
1 micromort. School shooting deaths are a very small part of
• the 30 000 shooting deaths each year;

• murders (the average murder risk in the United States is 50 micromorts
(see Table 4.1.1); or
• ‘natural’ deaths of school-age children as suggested by the UK figures in
Table 4.1.2. (One should check the US numbers, of course.)
The chance of having a shooting in ‘your school’ during the year 2022, a
traumatic event for the whole school and its community, was roughly 1 in
2000, averaged across the United States. Over the total period of schooling
of 10+ years, it is around 1 in 200. Whilst these risks are far too small to out-
weigh the benefits of school attendance for an individual, they are sufficiently
large to identify this as a problem of national significance.
Measures of likelihood
We conclude this agenda-setting section with a foretaste of the teaching chal-
lenges of reconciling the formal mathematics of probability with comprehen-
sibility in this area. The fundamental concept, probability, measured with
a number in the range 0–1, proves very difficult for people to grasp and,
particularly, to use. (No surprise for teachers there.) This is exacerbated by
the very small numbers often involved in risk estimates – 8 micromorts is a
probability of 0.000008. This difficulty is not confined to students in school;

there is a large body of evidence showing this, across a number of profes-
sions, including doctors (Hoffrage et al., 2000), particularly when compara-
tive probability is involved. Across media it regularly leads to eye-catching
headlines like “Research shows that … doubles the risk of developing cancer
of the”. But if the cancer is rare, then double the risk is also very low.
It is now well established that communicating about risk is most effective
when done in terms of numbers in a specific size of population; these are
sometimes known as ‘natural’ or ‘expected’ frequencies. Large whole num-
bers, though challenging to visualise, are more accessible than lengthy deci-
mals. For this example, if 1 person in a 1000 develops the cancer, doubling
the risk is not too serious; if the normal level is 100 in 1000, that is another
matter. Micromorts are, of course, the expected number of people dying from
a population of 1 million.
All this evidence strongly suggests that expressing probabilities as numbers
in a specified population is a good approach to developing mathematical and
statistical literacy. The classroom trials in the development of the online teach-
ing unit ‘How risky is life?’, described in the next section, reflected these results.
Appropriate precision
Risk is also a good area to bring out the important concept of appropriate
accuracy: that a calculated number cannot be more accurate than the as-
sumptions made and the numbers used in the calculation. For risk this often
means the numbers are only ‘ballpark estimates’ and more so because the
input data may be based on populations that are significantly different from
those where the conclusions will be used, which might often be just for an
individual. Mathematical literacy usually only requires rough numbers to
guide decisions – hence the numbers in Table 4.1.2 are rounded.
Appropriate degree of precision is important in all applications of math-
ematics (see Chapter 3), but it rarely gets much attention in traditional school
mathematics. It is not unusual to see nine digits, copied off a calculator as
an answer to a real-world problem, go unchallenged by the teacher. Lessons
focused on mathematical literacy are a good place to address this issue, b
ecause
knowledge of the context can provide clues to the accuracy of the data.
4.2 Exploring risk in the classroom

In this section we describe an approach to risk that tackles many of the issues
set out above in a five-lesson unit aimed primarily at 14–17-year-old students.
Designed and developed by the Shell Centre and reSolve teams, two versions
of the teaching materials are available free online [4B]; each gives detailed
teaching suggestions and supportive tools. The first, from Bowland Maths,
focuses on risk, while the second, from reSolve: Mathematics by Inquiry

(Australian Academy of Science), also addresses the modelling aspects more
explicitly. Feedback from classroom trials indicated that the learning and
enjoyment goals both were achieved by these sequences of lesson activities,
which we outline in this section. (Readers who are interested in design and
development may find it interesting to compare these two versions, developed
by overlapping teams for two different communities of teachers and students.)
Both “How risky is life?” units focus on a specific hazard (death) and
its likelihood, starting with deaths from unnatural causes – often a major
source of anxiety. Here we give an overview of the structure of these mod-
ules, reviewing the essential insights that each lesson addresses.
Impression versus reality

In preparation for the unit, the teacher encourages students to bring various
media reports of deaths into the first lesson, Exploring our perceptions of
risk. Groups of students rank their impressions of the probabilities of differ-
ent unnatural causes of death. The different groups’ views are then discussed
across the class and a consensus arrived at.
The second lesson, Sudden death – what does the data say? presents the
real data on the number of deaths per year for ten different causes, aiming
to bring out the mismatch between ‘common sense’ and reality in risk.
Three simple visual tools (designed by Malcolm Swan) are used to establish a
(logarithmic) scale of likelihood based on whole numbers, expressed as “1 in
10,000” etc., aiming to emphasise how small these risks are.
The first tool is a printed sheet with strips for each decade of likelihood,
described as “1 in …” and in words, where the students calculate the num-
bers of deaths in the whole population for each cause (using ballpark figures
of 25 million for Australia and 50 million for England). The second tool is a
poster that each group prepares using the real data and the likelihood strips,
sequenced in descending order of risk. Figure 4.2.1 shows a successfully com-
pleted UK poster. Note that the higher risk strips are not required. Again, the
numbers will be different for different periods – for example, the terrorism
number mainly reflects one incident.
The third device, also looking forward to the next lesson, helps students to
visualise the absolute size of each risk of death. It is simply a sheet of paper
with 100 squares in a 10 × 10 array, and then each such ‘large square’ is further
divided into 100 ‘small squares’. Students first calculate how many people in the
population each small square represents (5000 in the England case), assuming
the page represents the whole population. Then each student colours in an area
corresponding to each unnatural cause of death. In total less than two small
squares will be coloured, representing less than 200 micromorts per year – too
tiny to reproduce legibly here. The diagram can be found on the website [4C].
FIGURE 4.2.1 Likelihood of deaths from various unnatural causes (UK).

Source: From How risky is life? – Bowland Maths [4B].
FIGURE 4.2.2 The male–female differences in base risk among younger adults.
Source: From How risky is life? – Bowland Maths [4B].
Risk of death from all causes

The third lesson, The big picture, moves on to look at deaths from all causes –
now as a function of age. Graphs bring out the dominance of age in the scale
of risk across the lifespan. Other national data (supplied in the materials)
shows that poor health (i.e. a ‘natural’ cause) dominates, though unnatural
causes are about a third of the total up to age 40. The substantial differ-
ences between males and females from ages 15 to 30 shown in Figure 4.2.2
promotes discussion, as mentioned in Section 1.5.
ACTIVITY
Discuss the possible causes for the male–female differences shown in

the figure – and what further data you would look for to improve your
understanding.
Again, filling in the 10 × 10 array, total deaths fill roughly just one large
square – about 1% of the population die each year, overwhelmingly at older
ages as expected. So life is risky but death is very unlikely – until old age
where, if you get there, as we saw in Figure 4.1.2 you’re still likely to survive
a few more years.
The lesson goes on to present more detailed data on the way different
causes vary with age, involving translation between different representations.
It then goes on to probe understanding with a collection of statements to be
classified as ‘True, false, or impossible to tell?’ from the data provided.
Deeper issues of data literacy

The unit then goes on to explore four important data literacy concepts in a
little more depth. Centrally important is not taking data at face value but
critically probing the assumptions behind it. We then explore the mathemat-
ics of random variation, going on to discuss how far one can draw inferences
for individuals from population data, and the concept of ‘best estimate’.
Probing assumptions behind data

Going further, Lessons 4 and 5 of the reSolve unit explore these issues and
relate them to the process of data-driven modelling. Lesson 4 aims to help
students avoid just taking data at face value (even data from very reputable
sources), carefully considering what conclusions can really be drawn from it,
thinking critically about the situation and asking questions.
One activity is to decide if men are more dangerous drivers than women.
The data in Figure 4.2.3 is from the national statistics on cause of death
in Australia for 2015. This data clearly shows that at every age up to 85,
FIGURE 4.2.3 Number of deaths from road accidents in Australia in 2015 by age

and gender.
Source: Graph adapted from the reSolve How risky is life module [4B].
more men than women die in road accidents. Maybe men are more danger-
ous drivers than women, but they become more careful in very old age.
Students, working in groups to critically analyse the situation, will quickly
come up with reasons why the conclusion of dangerous driving cannot be
deduced from this data. The data includes road deaths for drivers, passen-
gers, and pedestrians. It does not account for how far or for how long men
on average were driving or were passengers in 2015 compared to women,
nor the possibility that men may more often have to drive in more dangerous
circumstances. Equally, change could be explained by there being far more
women than men over 85s, and hence more women over 85 driving. One
needs rates! We do not even know from this actual data that there was a sim-
ilar pattern in other years (although there is). Further information on many
additional variables has to be considered before we can decide whether men
are more dangerous drivers than women. Of course, in real life, it is very
likely that there will not be sufficient data available to make a decision on
any one specific question; and the best that can be done is to make reason-
able inferences from all available data, acknowledging limitations. This ac-
tivity is the heart of critical thinking, taking seriously the c ontext and all its
variables.
Random variation
reSolve Lesson 5 is about making predictions from data. It emphasises that
data (such as the number of road accident deaths) varies from year to year
because of random fluctuations and also because of ‘real’ changes that are
likely to persist. Simulation can help decide whether changes in data arise
from a change in the ‘real’ underlying cause or whether they are due to
random factors. A simulation app (a precursor to the study of confidence
intervals providing concrete examples) is used to show the magnitude of
variation in data that can be expected from these ‘random’ factors that have
no known significant meaning. The app can be used online or downloaded
from [4E].
For example, in 2017, the Australian Royal Life Saving National Drown-
ing Report [4D] reported 291 drowning deaths in aquatic locations across
Australia. That was an increase of 10 deaths (or 4%) from the rolling 10-
year average of 281 drowning deaths per annum. The simulation app can be
used to see if it is reasonable to conclude that the risk of death by drowning
really increased over the ten-year average. Figure 4.2.4 shows 20 simulations
of the number of drowning deaths assuming that the underlying probabil-
ity of dying this way is 281 in 25 000 000 (the approximate population of
Australia). In that case, 11 of the 20 simulations gave more than 290 deaths
per year. When the simulation was repeated 5 000 times, it showed that over
290 deaths would happen in about 40% of years from these random causes,
FIGURE 4.2.4
Simulated number of drowning deaths over 20 years assuming a
fixed risk.
Source: App by Shell Centre for Mathematical Education – used in Bowland Maths & reSolve
[4E].
without an actual change in the likelihood of drowning. The high number

of deaths (291) is disappointing when the desire is to reduce drowning, but
getting such a high number is easily able to happen by chance. There is no
evidence from this data that the underlying chance of a person drowning in
Australia had really increased in 2017. Understanding variation in data and
being able to put some measure on it is essential for interpreting data, and
hence for mathematical literacy. And a sad postscript: data since 2017 shows
steady increases.
From looking at multiple (simulated) examples, students are encouraged
to observe that n is usually a useful rough estimate of the random variation
to be expected in a situation where there is a rare event occurring about n
times in a given time period. (Here about 280 deaths per year; another ex-
ample is number of traffic accidents per year.) This rule (which comes from
the Poisson distribution) is an example of useful ‘know about’ mathematics
since most people never study the Poisson distribution. Compared with other
subjects, mathematics is reluctant to include ‘know about’ results, insightful
though they can be. For the 281 drownings, n is about 17, so 10 is well
within expected random fluctuations.
What can I infer about me?

The data use in lessons to this stage is ‘old’ data, and some from just one year.
Furthermore, this data comes from a large population that includes people
who “aren’t like me”. This raises two important general questions:
• How far does this data still apply now, several years later?
• How far does this data apply to me, an individual?
The aim of the guided exploration is that students come to see that this
data provides a best estimate from the data we have. More information,
involving more variables than just age and gender, could narrow the popu-
lation sampled and allow a better estimate, but there will be no data about
populations of people exactly like you.
The other big question is what would you do with the knowledge. Some
people won’t take any risk that they know about, while others do so all the time.
ACTIVITY
The PISA task EARTHQUAKE (M509 from the 2006 released items [4M])
assesses students’ basic ability to interpret probability statements (as do many
other PISA items).
4.3 Finding reliable data – critical thinking for data literacy

In this section we look more broadly at the challenges of data literacy – in
particular, how students can learn to find and sort through the data they need
from the flood of seemingly relevant data that is now available online. This
critical thinking is at the heart of data literacy. We explore it in the context of
risk, but the approach is relevant wherever data is needed, whether in data-
driven or theory-driven modelling approaches.
We start from the position that mathematical literate people have busy
lives. They do not want to devote large amounts of time to analysing things
that are not critical to their current decision-making. Time-effectiveness in
acquiring information is central. Where does one begin?
• How do you find information on an issue that interests you?

This is relatively straightforward. You start by putting the area of your
interest into a search engine, trying various wordings. You will be faced
with far too many links, often literally millions, starting with paid-for
‘Ad’ links (normally best avoided). This forces you to think about what
the important variables are. As you begin to formulate a model of the phe-
nomenon, you ask: ‘is this data a good measure for what we are interested
in, in this particular context’ – sometimes called ‘internal validity’.
• How do you probe how trustworthy the data is?
First, one is looking for reliable sources that have no inherent biases –
though no writing is value-free, some is more obviously ‘selling’ a point
of view. Wikipedia is usually a good place to start; it is monitored by
a community that feels strongly about objectivity, with clearly specified
sources. If in doubt, ‘digging deeper’ on Wikipedia will also reveal com-
munity discussion on the accuracy of the article.
Recently, ‘artificial intelligence’ systems, such as ChatGPT, have
become available to provide extended responses to questions posed in
plain language. At the time of writing, however, these sometimes produce
plausible sounding but completely fictitious answers. In one early case, a
lawyer used such a system to produce an impressive-looking legal argu-
ment, but it cited past cases that simply didn’t exist!
However reliable the source, two questions must always be asked.
What other influences might affect this data? What other variables should be
considered? The dangerous driving discussion above exemplifies these issues.
• What kinds of data are relatively straightforward to interpret?
This is at the heart of data literacy, and its relation to modelling. Ideally
one looks for data from a reliable source where only one variable is chang-
ing throughout the data set. This data can be compared to a model of the
underlying mechanisms, if you have one. Most data is, in contrast, ‘messy’
with several variables changing across the data set. Modelling such situa-
tions is thus more complex, usually involving a more heuristic approach,
with parameters determined from experiment.
• How far are the results generalisable to the situation I am interested in?
This question arises whenever, we want to make inferences from data
that comes from populations or situations that are different from the one
we are interested in. Is this generalisation warranted? Called ‘external
validity’, this can only be assessed by thinking critically through all the
other variables that might be significant in the phenomenon you are study-
ing and informally modelling the likely influence of each. Such critical
thinking is at the heart of mathematical literacy. In Section 4.4 we will
exemplify these general issues in the context of epidemics.
• Where to find data
National statistics agencies (such as the Singapore Department of Sta-
tistics) and organisations such as the World Bank, UNESCO, Gapminder
and even the CIA publish large data sets online and are usually highly
reliable. These can be used to explore a wide range of topical social and
economic issues, either with the online visualisation tools supplied by the
organisation or by downloading the raw data.
4.4 Epidemics – particularly COVID-19

The COVID-19 pandemic brought discussions of risk from academic jour-
nals aimed at experts onto the front pages of newspapers. Since serious
epidemics occur every decade or so, understanding the dynamics and the
meaning of the multiple kinds of data produced, and their implications for
the decisions that you as an individual make, are an important aspect of
mathematical literacy. We shall use this as a case study to illuminate some
general issues of data analysis and modelling and the challenges it presents
in complex situations that are changing rapidly. We look first at the fun-
damental dynamics of epidemics, moving on to some intriguing aspects of
COVID-19.
The epidemic curve

We begin with a standard simplified model of epidemics, which students can
explore using the spreadsheet [4F].
The basic effects are two: exponentially growing spread and increas-
ing immunity. Consider a closed population of people who interact with
each other. It is a reasonable model assumption, supported by evidence,
that when few people are infected, each infected person infects, on aver-
age, a given number, say R0 , other people. The infection will then grow
exponentially: if R0 = 3 , the number infected in each step of the develop-
ing infection is 1, 3, 9, 27,…. But it then follows that there is a growing
number of people who have been infected 1 + 3 + 9 + 27 + . Assuming that
a person cannot be infected twice (increasingly proving not to be the case
with COVID-19 and its variants), an increasing proportion of the people
an infected person interacts with will have become immune. Consequently
R, the average number of people infected by an infected person, decreases
from its initial value Ro . When R drops below 1, there is an exponential
decrease in the number of new infections and the epidemic dies out. The
population is said to have acquired herd immunity. This behaviour can
be explored by varying Ro in the spreadsheet. This behaviour is shown in
Figure 4.4.1.
ACTIVITY
Use the spreadsheet by varying the parameters to explore the level of infection.
Watch out for values of Ro that produce chaotic behaviour caused by step-by-
step ‘discrete mathematics’ effects.
FIGURE 4.4.1 Infection numbers in the classical model of an epidemic.

Source: Population of 1,000 people with Ro = 1.3 – see the spreadsheet at [4F].
Ro depends on many factors: the inherent transmissibility of the disease,

the time a person remains infectious, the average frequency of contact of
individuals, the time spent together, and other factors – all integrated across
the whole population at risk. All this makes calculating R from first princi-
ples complex, though simulation modelling gave useful results throughout
COVID. In practice, Ro was determined empirically from the speed of the
initial exponential growth of infections in the epidemic. Beyond its natural
contagiousness, Ro for COVID was increased by two features of the disease:
people were infectious for several days before they showed symptoms, and
some infectious people never developed symptoms at all.
This is an example of two complementary approaches to modelling
described in Chapter 1: theory-driven and data-driven. Theory-driven mod-
elling identifies the relationships between variables, but data-driven model-
ling works, for example, from graphs of empirical data – a mathematical
representation of a phenomenon with no underlying assumptions beyond
the choice of variables deemed to be of interest. For most situations a mix-
ture of approaches is appropriate. For epidemics we have a theory-driven
understanding of the initial exponential growth and the transition to herd
immunity but the values of the key parameters, Ro, and the timescale of each
‘step’ are to be found from observation in the initial stages of the epidemic.
How are the ‘steps’ in the transmission of the disease related to time?
Often through a parameter that quantifies the rate at which new cases are
arising, usually labelled r. This parameter r depends on R and the timescale
over which each step of infection occurs. A more accessible concept is

the doubling time of the epidemic td . With an exponential growth model
(number infected at time t = ke rt ) for the early stages before immunity devel-
ops, the number of infected people at time t = 0 is ke r ×0 = k. This will be
double at time td , where
k ⋅ e rtd ln ( 2) 0.7
= 2 so 2 = e rtd and td = or r =
k r td
Using specific units, a doubling time of a week corresponds to a rate
r = 0.1 per day.
COVID-19
That the COVID-19 pandemic was more complicated than described by this
simple model is made clear by the graphs of cases and deaths (those with
COVID-19 on death certificate) over time, shown for the UK in Figure 4.4.2.
ACTIVITY
Ask small groups to list factors that may have influenced the spread of COVID-19
and the numbers reported to produce the data shown in each of these graphs.
The data, particularly on case numbers, is messy. Indeed, the succession of

peaks shown in Figure 4.4.2 looks more like a sequence of different diseases.
The complicating factors include
• limited testing in the early stages when the real case numbers were vastly
greater,
• uneven patterns of reporting,
• virus mutations, and
• timing of countermeasures used to reduce transmission (masks, lock-
downs, etc.).
Restrictions were removed completely in February 2022.

The data on deaths is more reliable, but still affected by the countermeas-
ures, and the varying death rates associated with successive variants, miti-
gation by vaccination, and improving treatments. Reliable inferences about
mechanisms and risks needed more detailed and cleaner data than these.
Figure 4.4.3 shows data from the UK Office of National Statistics (ONS)
which measured the level of active infection by testing a random sample of
the at-home population in England, each week from July 2021 to June 2022.
FIGURE 4.4.2 Daily new reported cases and deaths from COVID-19 in the UK,
April 2020 to July 2022.
Source: Graphs adapted from coronavirus.data.gov.uk [4G].
FIGURE 4.4.3 The infection level in England over time for COVID-19.
Source: From weekly random samples – source data: ONS Infection survey July 2022 [4G].
It shows much the same qualitative features as the equivalent period in

Figure 4.4.2. This data allows a reasonable inference that COVID-19
remained highly infectious, with 1 person in 15–30 having the disease at any
one time in the first half of 2022 – as high as at any time in the pandemic.
DNA analysis of virus samples showed that successive peaks were asso-
ciated with increasingly infectious subvariants. Fortunately, with a now
largely vaccinated population, death rates remained relatively low in rich
countries.
This data alone makes clear that COVID-19, and indeed any real epi-
demic, brings in many other effects, so the analysis becomes more complex
than what is included in the simple epidemic model, with many more vari-
ables. No surprise, but what are these effects?
Diffusion
Rather than a closed interacting population, any infection starts in one place
and spreads – a complex diffusion process that depends on rates and patterns
of people mixing, locally, regionally, and internationally. The highly infec-
tious and lethal Black Death had spread from Mongolia into and through
Europe between 1347 and 1351, killing a third of the European population.
For the 1918–1920 virulent influenza epidemic, the origin is disputed, with
Kansas in the United States the leading contender, but it spread through-
out the United States and Europe in a series of waves over 3 years and was
further spread by movement of people at the end of World War I. (Though
widely known as “the Spanish flu”, Spain was certainly not the origin – just
the first place where the media were allowed to report it.) For recent epidem-
ics like COVID-19, modern air travel greatly speeds up this process. From its
first detection in China in late 2019, where it spread rapidly in the city of
Wuhan, the COVID-19 virus (properly named SARS-Cov-2) was carried by
Chinese skiers to northern Italy, quickly spreading so as to overwhelm the
regional hospital system in Italy in February 2020 by which time cases were
rising all over the world.
The physics of diffusion is well understood in simple situations like dis-
solving a solid in a still liquid or cigarette smoke in a room. The spread of
virus aerosols in carrying COVID between individuals was analysed at this
level, highlighting the importance of ventilation. But the diffusion of the epi-
demic at a macro-level involved many effects originating with people’s habits
of mixing at home, in travelling, and at work – for which the parameters
were rough estimates at best. The resultant modelling gave correspondingly
diverse predictions.
Virus mutation
All organisms mutate through random changes to their DNA; the simplest
ones like viruses change more quickly. The process of natural selection en-
sures that the variants that come to dominate are those that spread fastest
(i.e. have higher Ro). Killing people is a side effect that is, in fact, counterpro-
ductive from the ‘point of view’ of a virus that is transmitted when people
meet – dead people don’t move around. In some severe diseases, like Ebola
or SARS-1, people become seriously ill immediately, so Ro is low, determined
largely by the sanitary practices in nursing and burial. The COVID-19 virus
was thus a very ‘efficient’ virus, made more so through its rapid mutation
and, crucially, the limited cross-immunity between variants. Thus, the second
assumption of the simple epidemic model, that you wouldn’t catch COVID
twice, was not valid and herd immunity didn’t happen naturally.
Variable outcomes with age

For diseases like COVID-19 there are various levels of hazard: death, inten-
sive care, hospitalisation, more or less severe or long-lasting illness at home,
or even asymptomatic infection where the person is unaware that they have
the disease. For the individual the likelihood of different levels of hazard is
crucial for sensible decision-making. Again, we shall focus on risk of dying.
We have noted from Figure 4.1.2 that age is the central variable in determin-
ing base risk of death.
In Figure 4.4.4, the quantities graphed are base risk of dying and
‘COVID IFR’, the infection–fatality ratio, calculated here as the percent-
age of people who get COVID-19 who then die from it. Both graphs in
the figure show that, if you catch COVID-19, the age dependence of the
risk of dying is remarkably similar to the base risk. This is not always true
for epidemics; for example, the 1919 influenza epidemic was particularly
lethal for young adults. The absolute risk of a person dying after catch-
ing COVID-19 unvaccinated (at that time in the UK) roughly doubled for
every 8 years older.
But the figure also shows that the risk of dying if you had COVID
was about 60% of your base risk across all adult ages. In the language of
Figure 4.4.2, this is a 60% increase in your total micromorts per year, or
catching COVID-19 was equivalent to around 7 months (60% of a year) of
additional ‘normal risk’. So COVID-19 was indeed a serious disease, particu-
larly for older people. As always, more detailed data shows a more complex
picture with substantially different population mortality rates for different
sub-populations but the overall picture remains much the same.
FIGURE 4.4.4
Death rate with COVID-19 and base annual risk versus age on
linear (top) and logarithmic (bottom) scales.
Source: UK data through October 2020 (before vaccination became available). Combines data
from Brazeau et al. (2020) and ONS life tables.
ACTIVITY
a Discuss the advantages of linear versus logarithmic scales for showing vari-
ous kinds of functional behaviour – and why these different behaviours arise.
(Ans: Linear versus exponential – because the eye detects straight lines most
efficiently)
b Explain with examples how the conclusion that there is a 60% increase due
to COVID-19 was reached.
Mitigation measures
The measures taken by governments to slow the spread of COVID-19 covered
the whole range. It is true that the spread can be halted if every person is iso-
lated for the time from first infection to becoming non-infectious, about two
weeks for COVID-19; but the problems of sustaining life like this are obvious.
China nevertheless adopted an approach close to this, with very strict stay-
at-home measures; this kept case numbers low until the later more infectious
variants ‘broke through’. New Zealand used its island isolation to exclude all
incoming people, along with a rapid response to the few cases that got through.
In the UK there was a series of lockdowns as numbers started to rise – each too
late in the view of some experts – with relaxations in between. Influenced by
the early pictures from Italian hospitals, the key policy goal was to prevent the
health service being overwhelmed – “flattening the peak”. In the United States,
there was strong political pressure, which varied from state to state, against
using restrictions on ‘personal freedom’ – freedom to infect other people.
Modelling COVID-19
From early 2020 the COVID-19 pandemic led to an intense modelling effort
across many teams in many countries to predict the course of the epidemic
and to evaluate mitigation options. Ambitious early efforts at prediction pro-
duced wildly differing results, but it was then the best guidance available for
policymakers in making public health decisions that affected whole popula-
tions. In the UK, just before the first lockdown in March 2020, a consensus
of the various modelling groups led to an official statement that “20 000
deaths would be a good result”; in fact, total deaths from COVID-19 in the
following two years were over 200 000. A bad result.
As time went by and understanding of COVID-19 and its propagation
and treatment improved, the modelling became less ambitious and largely
descriptive, mainly with more sophisticated statistical analyses of recent data
along the lines of the simple epidemic model above, but applied to multiple
groups, in different places, with different R and r values. The COVID-19

experience underlines an important general result: that mathematical model-
ling of complex situations often gives only qualitative insights into the phe-
nomena involved. These can nonetheless be the best information available
for individual and community decision-making.
For COVID it was the vaccine research, remarkably accelerated by
methodological changes in the trial and approval process, along with im-
provements in treatment, that produced the breakthrough. Vaccination
turned a disease, which if left to run its course might well have increased
the normal death rate by more than 50%, into one that was relatively rarely
life-threatening. In fact, the rate of infection did not decrease for a long time,
driven by variants that were more and more infectious and it is likely to
remain significant – an endemic infection. While estimates of Ro for the origi-
nal ‘wild’ COVID-19 virus were around 4.5, estimates for Ro for the BA5
subvariant of Omicron that emerged in mid-2022 were initially around 18 –
matching that for measles (Guerra et al., 2017) along the benchmark for high
transmissibility. At least for the vaccinated, the disease caused was much less
likely to result in death than early in the pandemic. For further reading, see
the article “As COVID deaths in the UK pass the grim milestone of 200 000,
what have we learned?” from The Guardian [4H].
The overall picture

Perhaps the cleanest data on the effect of COVID-19 on the death rate is
the excess deaths over the average for the previous five years, shown by
Figure 4.4.5 for England. From April 2020 to June 2022, there were 1.26 mil-
lion registered deaths against 1.14 million ‘expected deaths’, which amounts
to a 10% increase, most of which could be attributed to COVID-19.
Implications for the individual

In what ways can this picture of the risks of COVID-19 help individuals
make better decisions? The data in Figure 4.4.4 shows each of us the abso-
lute risk of dying if you caught an early variant of COVID unvaccinated, and
how it relates to your base risk. It is clear that the increase in your base risk
is substantial, typically around 60%. As always, how you view this risk is a
personal decision. The equivalent data for fully vaccinated people makes it
clear that the risk of death (though not of infection) is very much lower. So,
unless you have other compelling reasons, you should get vaccinated or take
strong measures not to catch the disease. Even for the vaccinated, COVID-19
can be an unpleasant and sometimes dangerous disease, with symptoms that
persist for a long time in about 10% of those infected.
The other decisions are mainly about what mitigation measures you choose
to take: avoiding spending long periods in crowded indoor places, wearing a
FIGURE 4.4.5 Excess deaths in England during COVID-19 from March 2020 to

February 2023.
Source: Graph adapted from UK Office for Health Improvement and Disparities [4J].
mask, particularly when in such places, having your groceries delivered? For
such questions, finding and analysing the data provides a quantitative basis
on which the mathematically literate person can make informed decisions.
A readable analysis of the UK COVID data and our understanding of how
the epidemic developed is given in the book COVID by Numbers by Spiegel-
halter and Masters (2021).
4.5 Medical testing

Medical testing for infections and other ‘morbidities’ is an important part of
healthcare, in which improvements over recent decades have enabled earlier
treatments and improved outcomes for many serious diseases. But no test is
perfect – each test gives proportions of ‘false positive’ and ‘false negative’
results. This makes it an interesting and motivating context for developing
an understanding of conditional probability – a notoriously tricky topic –
in a risk-related context. This is partly because it produces some surprising
results; most of us find it surprising at first that even with a test that gives
a small proportion of false positives, most of the positive test results can be
false. For the mathematically interested person, surprising results like this are
a reason to look more deeply.
Teachers and their students take a variety of approaches to representing
the situation: using probabilities or calculating numbers within a given popu-
lation. They may use various representations: tree diagrams, tables, or even
Venn diagrams. In any approach, stating exactly what is being calculated is
crucial – and tricky: “the proportion of false positives” is ambiguous without
specifying the denominator. The standard use is the proportion of positive
test results that are false.
The Mathematics Assessment Project’s ‘classroom challenge’ on Repre-
senting Probabilities: Medical Testing [4K] provides a sequence of activities
that explore the problem in some depth. It starts with a specific problem,
comparing two countries A and B with different incidence of the disease,
20% and 2%, respectively. Samples of people from each country are tested
using the same test, for which trials have shown the following:
• If a person has the disease, then the test result will always be positive.
• If a person does not have the disease, then the probability of the test re-
porting that they have it, a ‘false positive’ test result, is 5%.
The question posed, for each country, is as follows:
ACTIVITY
Suppose a patient is told that they have tested positive. What is the probability
that the test is wrong?
The tree diagram representation for country A is shown in Figure 4.5.1. It

shows that the percentage of positive tests that are wrong is
0.04
= 17%
0.2 + 0.04
For Country B, with 2% incidence, the corresponding numbers are shown
in Figure 4.5.2. The 71% of positive tests that are false in this case arises
because the true positives drop in proportion to disease incidence; the false
positives increase only slightly. This is a nice example of how mathematical
literacy enables us to understand surprising results.
The spreadsheet shown in Figure 4.5.2 (see [4L]) enables students to
explore the effect of changing the input parameters. It also covers a test
that has a false negative rate as well. Changing the population entry from 1
translates probabilities into numbers in a population throughout the calcu-
lation. Perhaps most usefully, the ‘Show formulas’ mode of the spreadsheet
brings out the detailed reasoning, which many students find difficult to
construct.
In the MAP unit the students individually try the problem, which has some
guidance from specific questions, in a previous lesson. A hint sheet is pro-
vided for the teacher to use where needed – differentiation through support,
FIGURE 4.5.1
Probabilities for Country A with 20% incidence and a 5% false
positive test.
Source: Adapted from Representing Probabilities: Medical Testing [4K].
FIGURE 4.5.2 Probabilities for Country B with 2% incidence and a 5% false posi-

tive test.
Source: Spreadsheet available at [4L].
not in the core task, is a design principle. Looking through the responses
gives the teacher a base for formative assessment.
In the main lesson, the teacher clarifies the definition of false positives,
supported by an applet. Students share their attempts in small groups, de-
ciding on a common approach. The focus is on the chains of reasoning. In
the diagnostic teaching approach to formative assessment, a key design tac-
tic uses examples of “student work” to stimulate discussion of various ap-
proaches, in groups and later across the class. The examples are chosen to
show sources of error in reasoning that often arise. Figure 4.5.3 shows three.
FIGURE 4.5.3 Three student responses to the Country A ‘false positives’ task for
students to critique.
Source: From Representing Probabilities: Medical Testing [4K].
These use the idea of ‘expected frequencies’, for example, what does it mean
for 1,000 people being tested? This is clearer to most people than the (equiva-
lent) probability tree in Figures 4.5.1 and 4.5.2.
Amy has an error in her table – there should be 200 people, not 160, who
have the disease and test positive – but her method, for the proportion of positive
tests that are false, is correct. Noreen and Rajeev calculate a different quantity –
the proportion of false positives in the whole population. Rajeev’s calculation
is correct, and well explained. Noreen has an error in her Venn diagram – there
should be 760 people, not 800, who do not have the disease and test negative.
This device leads to focused discussion, moving students into the ‘teacher
roles’ of critiquing and explaining – a role shift that, as explained in Chapter 2,
raises the level of discussion from answers to reasoning. No arithmetic e rrors
are involved – trials showed that they divert attention from the reasoning.
Acknowledgements
Our thanks to David Spiegelhalter for educative and entertaining conversa-
tions on risk over many years.
Much of the data used in this chapter comes from UK government
sources – such as the Office for National Statistics and the Office for Health
Improvement and Disparities – and is public sector information licensed un-
der the Open Government Licence v3.0. See the individual attributions and
links for more information.
Material from the two versions of the How risky is life materials appears
courtesy of the Bowland Maths maintainers and the Australian Academy
of Science. Specifically, Figures 4.2.1 and 4.2.2 are from Bowland Maths ©
Bowland Charitable Trust 2008. Bowland Maths materials are free for edu-
cational use. Figure 4.2.3 is from reSolve: Mathematics by Inquiry © Austral-
ian Academy of Science, available under Creative Commons CC-BY-NC-SA
licence. Similar material appears in both versions. The software shown in
Figure 4.2.4, which features in both Bowland and reSolve, is from the Shell
Centre for Mathematical Education and available under Creative Commons
CC-BY-NC-SA licence.
Figures 4.5.1 and 4.5.3 adapted here with permission from the Bell
Burkhardt Daro Shell Centre (BBDSC) Trust on behalf of the Mathematics
Assessment Project. © Mathematics Assessment Resource Service 2007–2015.
The Classroom Challenges materials are available under the Creative
Commons BY-NC-ND 3.0 licence.
References
Brazeau, N., Verity, R., & Jenks, S. et al. (2020). COVID-19 infection fatal-
ity ratio: Estimates from seroprevalence. Imperial College London. https://doi.
org/10.25561/83545
Guerra, F. M., Bolotin, S., & Lim, G. et. al. (2017). The basic reproduction num-
ber (R0) of measles: A systematic review. Lancet 17(12). https://doi.org/10.1016/
S1473-3099(17)30307-9
Hoffrage, U., Lindsey, S., Hertwig, R., & Gigerenzer, G. (2000). Communicat-
ing statistical information. Science 290, 2261–2262. https://doi.org/10.1126/
science.290.5500.2261
Howard, R. A. (1980). On making life and death decisions. In J. Richard, C. Schwing,
& W. A. Albers (Eds.), Societal risk assessment: How safe is safe enough? General
Motors Research Laboratories. Plenum Press.
Spiegelhalter, D., & Masters, A. (2021). COVID by numbers: Making sense of the
pandemic with data. Penguin UK.

for example, ltml.mathlit.org/4A
purposes:
[4A] National Life Tables – Office for National Statistics (UK)

https://www.ons.gov.uk/peoplepopulationandcommunity/births
deathsandmarriages/lifeexpectancies/datasets/nationallifetables
unitedkingdomreferencetables
[4B] How Risky Is Life? – Bowland Maths & reSolve
https://www.bowlandmaths.org /projects/how_risky_is_life.html
https://www.resolve.edu.au/mathematical-modelling-how-risky-
life?lesson=1672
[4C] Visualising Small Risks – Bowland Maths & reSolve
https://ltml.mathlit.org/4C
[4D] Drowning Reports –Royal Life Saving Australia
https://www.royallifesaving.com.au/research-and-policy/drowning-
research/national-drowning-reports
[4E] Simulating Random Variations – Shell Centre for Mathematics
Education
https://ltml.mathshell.org/4E
[4F] Basic Epidemic Model spreadsheet
https://ltml.mathlit.org/4F
[4G] UK COVID-19 data sources – UK government & UK Office for
National Statistics
https://coronavirus.data.gov.uk/
https://www.ons.gov.uk/peoplepopulationandcommunity/health
andsocialcare/conditionsanddiseases
[4H] As COVID Deaths in the UK Pass the Grim Milestone of 200,000,

What Have We Learned? – The Guardian
https://www.theguardian.com/commentisfree/2022/jul/13/as-covid-
deaths-in-the-uk-surpass-the-grim-milestone-of-200000-what-have-
we-learned
[4J] Excess Mortality in England – Office for Health Improvement and
Disparities
https://www.gov.uk/government/publications/excess-mortality-in-
england-weekly-reports
[4K] Representing Probabilities: Medical Testing – Mathematics Assess-
ment Project
https://www.map.mathshell.org/lessons.php?unit=9405&collection=8
[4L] Medical Testing spreadsheet
https://ltml.mathlit.org/4L
[4M] Released Mathematics Items– PISA/OECD
Test questions 2006 https://www.oecd.org/pisa/testquestions-pisa
2006.htmTest questions 2003 – 2012 https://www.oecd.org/pisa/
pisaproducts/pisa-test-questions.htm
Direct link: https://www.oecd.org/pisa/test/PISA%202012%20items
%20for%20release_ENGLISH.pdf
5
CLIMATE CHANGE
The science, the mathematics,
and the politics
In the long term there is no greater risk than climate change, or other
topic as important – hence this separate chapter where, as usual, we hope
to set out some of the basic underlying phenomena, physical and societal,
in a way that stimulates constructive discussions between students on a
subject that is of interest and concern to so many of them. Because of the
complexity of these systems, this is mainly a ‘know about’ chapter in a
field where modelling and data complement each other – but the models
are complex.
The effects of human activity on the climate, predicted for over 50 years,
are already with us – in droughts, disappearing glaciers, and the worldwide
increase in extreme weather events, to name just a few examples. The dynam-
ics of the climate and its impact on weather are complex but the basic science
is clear, as is the kind of changes in human behaviour needed to mitigate the
effects of the heating.
In this chapter we look at how mathematical literacy can help each of
us to better understand the evidence on climate change, and the underly-
ing dynamics. We go on to look at the societal challenges and the actions,
personal and political, that will mitigate the damage that is ongoing – and
possibly even reverse some of it. We look at data, the pattern of arguments
that are deployed, some of the models being presented and the questions
they raise.
After first looking at the science we go on to describe the political and so-
cial frameworks within which climate change has been discussed since it was
first widely recognised by scientists as a potential problem in the 1970s. This
is key to relating an apparently straightforward set of scientific results to the
range of human responses to these results.
DOI: 10.4324/9781003303503-6
Climate change 107
Although relatively modest changes in climate can cause – and are already
causing – problems, a greater concern is that continued inaction could cre-
ate a self-reinforcing cycle of escalating change with even more catastrophic
results. So we look in more detail at the mathematics of feedback and tipping
points which makes such concerns credible.
Finally, we point to the parallel linked challenges of understanding the
much more complex effects of human activity on the ecosystem of the planet
and the methodological and ethical issues they raise.
5.1 The science of global warming

Variations in climate have been happening for millions of years, some
substantial ones within timescales of tens of thousands of years. The last
Ice Age lasted about 100 000 years, ending about 10 000 years ago. The
global average temperature was about 5°C lower than today. Larger varia-
tions happen in specific places from year to year and day to day – these we
call weather. The current focus is on changes with a timescale measured
in decades and how far these are anthropogenic, that is caused by human
activity.
An awareness that changes in atmospheric carbon dioxide levels could
substantially alter the surface temperature through ‘the greenhouse effect’
goes back over 100 years to a paper by the Swedish scientist Svante Arrhe-
nius (1896). In 1938 Guy Callendar connected carbon dioxide increases in
Earth’s atmosphere to observed global warming. By the 1960s internal CIA
reports recognised the phenomena and began to estimate its effects, societal
as well as physical. But the 1970s was the starting point of serious political
interest. In 1972 John Sawyer, a scientist at the UK Meteorological Office,
reviewed the scientific knowledge at the time on the likely effects of anthro-
pogenic carbon dioxide as a ‘greenhouse gas’, its distribution and rise. His
article (Sawyer, 1972) stated:
The increase of 25% CO2 expected by the end of the century therefore
corresponds to an increase of 0.6°C in the world temperature – an amount
somewhat greater than the climatic variation of recent centuries.
(p. 25)
In the outcome Sawyer rather accurately predicted the rate of global

warming for the period between 1972 and 2000, shown by the subse-
quent measurements from three national centres, such as the NASA data
in Figure 5.1.1.
This data begins in the nineteenth century, when thermometers were
introduced into meteorology. But we can now measure temperatures, and
many other variables, from thousands of years ago. How? One answer
108 Climate change
FIGURE 5.1.1 World average mean annual temperature relative to the average for
1951–1980.
Data source: NASA’s Goddard Institute for Space Studies (GISS) [5A].
lies in an ingenious bit of science and mathematics involving ‘ice cores’.

These are vertical columns of ice cut from the Antarctic ice covering, which
is formed from snow that has fallen over centuries, each layer compressing
those below. Some cores go up to 3 km deep representing 800 000 years
of atmospheric history. The physical properties of the ice and of material
trapped in it tell us about the climate over the age range of the core. The
proportions of different oxygen and hydrogen isotopes provide information
about ancient temperatures, and the air trapped in tiny bubbles can be ana-
lysed to determine the level of atmospheric gases such as carbon dioxide,
where carbon dating (discussed below) is a powerful tool.
The data from both thermometers and ice cores shows that the mean
temperature at the Earth’s surface has increased by about 1.2°C in the last
150 years – essentially since the industrial revolution literally ‘gathered
steam’. There are some other sources of data on past climate. On a geological
timescale, the pattern of life revealed in the fossil record gives some indication.
Over the last several hundred years, tree ring sequences give information on
local temperature and rainfall.
5.2 The mechanisms of global heating

Let us begin with some basic scientific facts:
• Carbon dioxide (CO2) is produced in great quantities by burning current or

fossilised organic matter that is rich in carbon: wood, coal, oil, or ‘natural’
gas. The atmosphere provides the oxygen.
Climate change 109
• CO2 traps the energy of sunlight within the atmosphere, as the glass traps
it in a greenhouse. The visible light of the sun passes through the ‘glass’
and warms the earth; the low temperature infrared light emitted by the
warmed earth (invisible to the eye, but you can sometimes feel it) cannot
penetrate the atmosphere to escape back into space. Some other gases also
contribute to the ‘glass’.
• CO2 is inert and stays in the atmosphere for a long time, with a ‘half-
life’ of about 120 years – that is, half of what is present disappears each
120 years.
• Natural processes create CO2 – animals breathing out, for example – but
this is part of the balanced ‘carbon cycle’ in which plants absorb CO2 from
the air, animals feed (directly or indirectly) on those plants and release
some of that recently absorbed CO2 back into the air. CO2 released by
burning fossil fuels, on the other hand, comes from carbon that has been
locked away from the environment for millions of years, and is the main
cause of increasing CO2 levels.
That CO2 absorbs the infrared (thus preventing it radiating into space) is
straightforward to demonstrate. Figure 5.2.1 shows one experiment (Leven-
dis et al., 2020) [5B]. A heating element at 50°C is hung in the middle of a
‘bouncy ball’ balloon filled with either CO2 or air. After being turned off, the
heating element cools more slowly with the CO2, an effect that increases with
increasing amount of CO2.
FIGURE 5.2.1
Schematic of an experiment to show the infrared absorption
by CO2.
Source: From Levendis et al. (2020) [5B].
110 Climate change
FIGURE 5.2.2 Concentration of CO2 in the atmosphere— over last 1000 years.
Data source: MacFarling Meure et al. (2006) via NOAA.
That burning carbon produces CO2 is well known (see, for example, [5C]).
One tonne of coal produces about 2.4 tonnes of CO2; not quite the ratio you
would expect from the molecular weights (C = 12, O = 16, CO2 = 44) because
coal is not just carbon.
The atmospheric effect of burning carbon in its various forms is shown
in Figure 5.2.2 which shows the parts per million of CO2 in the atmosphere
over the last 1000 years. The line shows data from recent direct measure-
ments in the atmosphere. The blobs are data from ice cores collected by the
Australian Scientific Expedition to the Antarctic, from the Law Dome which
rises to over 1900 meters above sea level. (Note: ‘Firn’ refers to a type of ice
that is part-way between snow and glacial ice.)
Each layer of ice in the core can be dated by measuring the proportion of
the radioactive isotope 14C that remains. How does this ‘carbon dating’
work? 14C is produced in the atmosphere mainly by cosmic neutrons and,
because it is chemically the same as 12C, it is absorbed by plants in the usual
way. This produces about 1 part in 1012 of 14C in the atmosphere. This decays
exponentially with a half-life of 5700 years to make 12C, and emitting radia-
tion. Measuring the radioactivity being emitted from a layer in the ice core
indicates how much of the 14C is left, and thus dates the layer.
ACTIVITY
Find other uses of carbon dating and how they work – in archaeology and in
other fields. How does carbon dating link to tree ring dating?
Climate change 111
Further confirmatory evidence from ice cores showed that the proportion of
CO2 and temperature had gone up and down together in wide swings through
past ice ages. This confirmed the planet-scale CO2–temperature relationship
in a manner entirely independent of computer climate models, strongly rein-
forcing the emerging scientific consensus. The findings also pointed to power-
ful biological and geochemical feedbacks – we discuss feedback mechanisms
below.
Looking more deeply

There are other gases that contribute to the greenhouse effect but they have
far shorter half-lives, so they disappear more quickly:
• Methane, which is released from oil wells, fracking, and the burping of
cows, is 80 times more absorbing than CO2 but with a half-life of ‘only’
10.5 years.
• CFCs, chlorofluorocarbon molecules which have been used in refrigera-
tors, have half-lives from 16 to more than 500 years. These were mainly
responsible for the ‘hole’ in the Antarctic ozone layer that helps protect us
from the ultraviolet in sunlight. Thanks to a worldwide effort to eliminate
the most damaging CFCs, the ozone layer has recovered.
• Nitrous oxide, which has a half-life of 132 years, also contributed to
ozone layer depletion.
But none of these other gases are emitted in anything like the same
quantity as CO2. Global carbon dioxide emissions from fossil fuels and
other industrial processes are at record high levels of about 40 billion
metric tonnes (GtCO2) per year, about 5 tonnes per person on Earth – the
US and European average is several times that. About 2% of this is pro-
duced by humans breathing out – about 100 kg per person per year. To
measure the differential effects of the various greenhouse gases, the IPCC
introduced the concept of ‘global warming potential’, comparing the ef-
fects of equal masses of greenhouse gases over different periods of time,
taking into account their half-lives and infrared absorption. CO2 is used
as the reference gas, so its GWP is set at 1 for any time period. Some 100-
year GWPs are shown in Table 5.2.3. Most gases with very high GWP
have only a low concentration in the atmosphere, so they are not the main
factor contributing to climate change. Finally, there is water vapour, a
most powerful greenhouse gas because of its infrared absorption. Ironi-
cally, it has negligible GWP because human-generated water vapour (e.g.
from power station cooling towers) leaves the atmosphere within a week
or two. However, it plays a crucial role in the feedback mechanisms that
we discuss below.
112 Climate change
TABLE 5.2.3 Global warming potential of some greenhouse gases
Greenhouse gas Chemical formula Half-life 100-year GWP

years
Carbon dioxide CO2 5700 1

Methane CH4 12 25
Nitrous oxide N2O 109 298
Water vapour H2O ~0 ~0
Some fluorocarbons – refrigerator gases
HFC-23 CHF3 222 14 800
Difluoromethane (HFC-32) CH2F2 5 675
Fluoromethane (HFC-41) CH3F 2 92
Source: Compiled using data from Wikipedia [5D].
ACTIVITY
Using this and other data as necessary, estimate the 25-year and 500-year
GWPs of these gases.
5.3 The politics of climate change

The 1970s was the starting point of serious political interest in the effect of
CO2 emissions on the climate. It perhaps began with John Sawyer’s 1972
prediction of an increase of 0.6°C in the world temperature by 2000. By
1977 Frank Press, US President Carter’s Science Adviser, had sent the presi-
dent a memo stating:
Because of the “greenhouse effect” of atmospheric CO2 the increased con-

centration will induce a global climatic warming of anywhere from 0.5 to
5°C. … The potential effect on the environment of a climatic fluctuation
of such rapidity could be catastrophic and calls for an impact assessment
of unprecedented importance and difficulty. A rapid climatic change may
result in large scale crop failures at a time when an increased world popula-
tion taxes agriculture to the limits of productivity. … The urgency of the
problem derives from our inability to shift rapidly to non-fossil fuel sources
once the climatic effects become evident not long after the year 2000.
Around the same time, internal research at the US oil companies came to
much the same conclusions. Supran et al. (2023) [5E] show that one of those
fossil fuel companies, ExxonMobil, had their own internal models that pro-
jected warming trajectories. Figure 5.3.1 compares Exxon’s predictions with
Climate change 113
FIGURE 5.3.1 Atmospheric concentration of CO2 – predicted versus measured.

Source: Data from ExxonMobil 1980s predictions (Supran et al., 2023) [5E] and NOAA meas-
urements [5F].
modern measurements from NOAA/NASA. Their predictions were consist-

ent with those forecast by the independent academic and government models,
and were confirmed by subsequent observation of atmospheric CO2 (NOAA
via NASA, [5F]).
Yet for decades members of the fossil fuel industry have worked to con-
vince the public that anthropogenic climate warming is not real or, at least,
not serious. What they knew about climate change thus contradicted what
they led the public to believe. The obvious commercial incentive to minimise
changes in the balance of different sources of energy is pursued through a
variety of strategies like those developed by the tobacco industry in the 1950s,
as evidence grew that smoking causes lung cancer. They argue that the models
used are too uncertain to show a causative link between the increase in CO2
from fossil fuel use and climate, and that the clear correlation does not imply
causation. Yet, for climate change as for cancer, the predictions of the models
have proved remarkably accurate over half a century and causation has been
established through identifying the mechanisms underlying the observations.
Prediction and then testing against subsequent events, and understanding
mechanisms are fundamental in establishing scientific validity.
Gradually over time there has been a shift in the defensive arguments
towards accepting global warming is a fact but denying its seriousness and
so the need to do much about it. This is often combined with ‘greenwashing’ –
publicity from oil companies and others that aim to improve their image by
describing steps to reduce carbon emissions that exaggerate their likely effect.
114 Climate change
For example, ‘carbon offsetting’, where trees are planted that will, over dec-
ades, absorb as much CO2 as was emitted in the flight they just took yester-
day. There can never be enough trees in the world to offset current carbon
emissions; there is no substitute for cutting emissions.
ACTIVITY
Justify the statement that there can never be enough trees in the world to offset
current carbon emissions.
The ‘debate’ continues to shift. The overwhelming majority of scientific pa-

pers (literature reviews find 97% or more) provide evidence of anthropogenic
causes of climate change and its potentially serious consequences: sea level
rises causing floods of coastal land, some places becoming uninhabitable
through temperature, food crop failures, and mass migrations. The fossil fuel
companies continue to construct specious arguments, finding – and funding –
scientists who support them. They are not alone. There are also strong strands
of political and public opinion that seek to undermine both the scientific facts
and, particularly, the need for making challenging societal changes. Unfor-
tunately, the current tendency to polarisation of political beliefs is making
resolution through rational argument based on data more difficult.
There are two main dimensions in the way governments and their citizens
have responded: danger level and action planned. The range of response in
each is broad, according to the assessment of risk.
Danger level:
1 Hoax: humans have no significant influence on climate.

2 Minimal effects: there are some effects but nothing to worry about.
3 Serious challenges: this is a problem and we must adjust our lives to ad-
dress it.
4 Potentially catastrophic: this is an upcoming disaster which must be top
priority in policymaking and in individual choices.
Action required:
a None.
b Awareness, along with some personal efforts to reduce our CO2 profile.
c Modest political adjustment, with policies that claim to meet international
targets – but, given how far rhetoric is ahead of action, they are unlikely
to do so.
d Steps that amount to a major change now in the way we live.
Climate change 115
Although we lack data, it seems likely that in many Western countries

there is a significant proportion of 1a people, a majority in 3b or 3c, and a
small proportion of really concerned 4d – along with many ‘don’t know’s. As
we write most governments seem to take a 3c position, 3d or 4d being politi-
cally challenging for all governments:
• For democracies because of the need to keep voters on side.

• For autocratic regimes because of a general priority for economic growth.
The mass of data available for thinking through these issues presents
both an opportunity and a challenge in trying to come to a reasoned
position.
ACTIVITY
Working in small groups, gather and analyse data to prepare an argument

for a debate on what actions should be taken to reduce global warming,
and to mitigate the effects of climate change. Each group will take a spe-
cific role, for example, fossil fuel company publicist, climate change activ-
ist, conservative or progressive politicians, leader of a Pacific Island, marine
biologist, world food or human rights advocate. (Draw lots if necessary to
ensure coverage.)
5.4 The mathematics of feedback and tipping points

Because we have been setting out something of the complexities of climate
change, and the scientific understanding of it, most of the mathematical liter-
acy demands have been in been understanding the meaning of data, presented
in numbers or graphs, and the units and rates involved and proportional rea-
soning about its implications.
We conclude this chapter with the mathematics of two dynamical mecha-
nisms that can make change catastrophic: feedback and tipping points. They
are found in many contexts, including aspects of climate change. Both can
be modelled within standard mathematics in ways that vividly illustrate the
general principles.
Feedback
Feedback (strictly ‘positive feedback’) is the term used when a change leads
to more of the same change. For example, the white ice in the Arctic Ocean
116 Climate change
reflects a lot of sunlight back into space, whereas open water absorbs it and
warms up. This creates a positive feedback loop:
1 Temperatures of ocean and atmosphere increase slightly (for any reason).

2 → There is less ice and more open water.
3 → Less sunlight gets reflected and more is absorbed by the ocean.
4 → The temperature of the ocean (and then the atmosphere) increases further.
5 Repeat from step 2!
To take a more mathematical approach, say that the ice cover in the Arctic
Ocean in the winter lasts for a period T each year. If the warming effect comes
largely from the planet as a whole – that is, there is no “feedback” from local
changes – and we assume it is steady over time, there will be a linear decline
in T from the starting year T0 to the year when T = 0. (Why does the model
stop working when T = 0?)
If, however, the local warming effect is also important, then the rate of
reduction will increase as T gets shorter, so the solutions to the new models
show exponential behaviour, characteristic of feedback. It cannot go on for-
ever, of course, as other mechanisms intervene.
dT
= − a (T0 − T ) where t is the elapsed time and a is assumed constant
dt
− ln (T0 − T ) = − at + b solving the differential equation, with b a constant
T0 − T = c.e at
where constant b is − ln ( c )
T = T0 − c.e at the ice cover reduces exponentially over time
In reality, both local and global effects are present which lead to a more
complicated model, but still expressible and solvable by calculus. As usual,
when the model is further improved by taking other factors into account,
numerical methods are required. Numerical methods, step by step using a
spreadsheet, can be used to demonstrate the effect to pre-calculus students.
There are many feedback effects in the dynamics of climate. The most
important ever-present one is more complex: the effect of water vapour. This
arises because the amount of water that the atmosphere can hold increases
rapidly with temperature. This drives the observed and predicted increase in
storms and other extreme weather events. Water vapour amplifies the effect of
CO2; without it, even doubling the pre-industrial level of CO2 (to 560 ppm)
would lead to an increase in global temperature of only about 1°C. The sea
is also a principal repository of CO2; the amount it can hold decreases with
rising temperature, creating another positive feedback mechanism.
Feedback mechanisms make modelling the climate with mathematics very
challenging. Although the many simulations of the diverse modelling groups
vary a lot in detail, the overall picture is consistent.
Climate change 117
Tipping points – obvious and surprising

A tipping point arises when an infinitesimal change in the parameters of a
system produces a finite shift in the state of the system.
Tipping points are a familiar phenomenon. In the literal sense of ‘tip-
ping point’, a tall object will topple over if it is tilted just a tiny bit beyond
a certain limit. Generalising the concept, when an election is lost by even a
few votes, a government can change. A kettle boils when the temperature
reaches 100°C – then it stays there, all the extra energy making steam. The
popularity of soccer is said to arise from the tension that arises because each
goal is seen as a potential tipping point in the match. The claim is that the
game is fascinating because there are so few goals not as some might think,
despite that.
How might tipping points affect climate change? There are various sys-
tems of concern. The ground in a huge area of Siberia is frozen throughout
the year. This permafrost holds a large amount of CO2 which will be released
into the atmosphere if the ground temperature moves above 0°C. The tem-
perature is rising fast in Siberia, hence the concern about this tipping point.
The Atlantic Meridional Overturning Circulation (AMOC) is a system of
ocean currents, including the Gulf Stream, that redistributes heat by bringing
it from the tropics all the way to Greenland and carrying cold water back
to the equator; it is weakening and, if it reaches a possible tipping point, it
could bring back Ice Age conditions to Europe. A very surprising outcome of
global warming.
How does mathematics model tipping points? Let’s start with some literal
tipping points. When a ladder leans against a wall, there are at least six ways
it can tip over; two are sketched in Figure 5.4.1.
FIGURE 5.4.1 Two potential tipping points for ladders.

118 Climate change
The “too steep” instability shown in the Figure is obvious, and benign.
As you step onto the ladder, your weight, acting through your hands, pulls
the ladder backwards. It begins to tip away from the wall, so you step off
and set it at a less steep angle. The “too flat” case has long been used as a
standard high school mechanics task – if the angle of the ladder is too flat,
the bottom slips away because the friction at A is inadequate. The student
has to relate this critical angle θ to the coefficient of friction. Ironically, this
is the most dangerous of many possible instabilities because it gets worse as
the climber goes up the ladder; once the climber is past the centre of the lad-
der, the ladder is more likely to slip than before – but there is no climber in
the traditional examination task! In the real world, this instability explains
the traditional safety rule about having another person standing on the bot-
tom of the ladder throughout, so the centre of gravity of the pair remains
below that of the ladder itself – provided they are of comparable weight, not
Laurel and Hardy.
The mathematical analysis of tipping points was called Catastrophe The-
ory by René Thom, who won a Fields Medal (“The equivalent in Mathemat-
ics of a Nobel Prize”) for classifying topological features of surfaces that
produce tipping points (Thom, 1977). We can get the idea from a simple
one-dimensional case shown in Figure 5.4.2.
When P = 0.5, the ‘ball’ in Figure 5.4.2 is sitting ‘safely’ in the dip of this
helter-skelter cubic curve – the nearby point of lowest gravitational potential.
Even if the ball is displaced slightly, it will roll back to the stable position.
The other curves show the graphs for various values of P. Note how the dip
gets shallower as P increases, disappearing at P = 4/3 – the tipping point when,
FIGURE 5.4.2 (
Graphs of the potential function y = x P + ( x − 2) .
2
)
Climate change 119
mathematically, the two roots for dy /dx = 0 coincide, so the maximum and
minimum merge into a point of inflexion. There the curve is completely flat.
Physically, with no local minimum potential, there is nothing keeping the ball
in position, so the slightest displacement from the inflection point will send
it rolling away down the slope. With P > 4/3, there is no flat area at all for
the ball to rest.
Even with P = 0.5, push the ball too far to the left and it will roll down the
slope. As P approaches the ‘tipping point’ of 4/3, the ‘safety margin’ – how
far the ball can be moved and still return to the equilibrium point – becomes
smaller. This is an example of a ‘metastable equilibrium’ which is a common
feature of the cycles and balances on which our environment depends. Arti-
ficially pushing the environment close to a tipping point reduces its resilience
to other, unavoidable disturbances with natural causes.
The phenomena described in this chapter – physical, mathematical, and
societal – underpin the need for critical thinking that leads to far more
effective action than the start that societies around the world have made
so far.
5.5 Climate change and the environment

We cannot leave the issue of human activity and climate change without re-
ferring to their direct consequences for the ecosystem in which we live. The
environmental effects of climate change are even more complex and diverse
than the physical changes we have discussed so far – and correspondingly
more difficult to model. Sorting and understanding the data alone is a chal-
lenge, even before tangling with issues of causality. Here we shall outline
some smaller scale phenomena which illustrate these challenges before going
on to discuss some theoretical frameworks people have put forward to think
about ecological impacts.
Climate versus environment – some examples of complex trade-offs

Efforts to reduce climate change can bring their own environmental con-
cerns. The following are examples of the unintended consequences of well-
intentioned initiatives. They illustrate the difficulty of designing interventions
in complex interacting systems, and the ethical values so often involved.
In the early years of this century, the better fuel consumption of diesel
engines, compared to gasoline engines of the same power, led to a boom in
diesel cars and government action to promote them. They are cheaper to run
and have lower CO2 emissions – better for both the purse and the planet.
“What’s not to like?” However, it was then recognised that diesels emit more
of the fine particles and toxic gases that are particularly harmful for the
lungs – worse for public health, particularly in cities.
120 Climate change
Most discussion of ‘climate change’ is, like ours above, primarily con-
cerned with the effect of reintroducing long-buried CO2 to the atmosphere
by burning fossil fuels. This is made worse by permanently clearing large
areas of long-established forest, which remove significant quantities of
CO2 from the air, perhaps to produce food or fuel. Burning even sustain-
ably produced ‘biofuels’ releases CO2 that was only recently absorbed
by plants or trees, expecting it will be reabsorbed when those plants are
regrown. This is ‘good’ for climate change, but there are downsides for
the environment: heating your house by burning wood produces smoke
laden with dangerous particles and toxins – a problem in densely popu-
lated areas.
Many types of plant material contain oils or sugars that can be fermented
to alcohol, providing good, clean fuel for cars, with no fossil CO2 emissions
– unless, of course, government subsidies on growing those plants encourage
farmers to cut down forests or stop growing food crops, raising food prices.
The Bowland Maths project You Reckon? [5G] includes a task which asks
students to investigate the pros and cons of this.
Nuclear power has, traditionally, been anathema to the conservation
movement. The long-term problem of dealing with spent fuel and decommis-
sioned reactors and the small risk of potentially catastrophic accidents are a
grave concern. However, nuclear fission does not release greenhouse gases,
and nuclear fuel can generate much more energy than a comparable amount
of conventional fuel [5H]. Moreover, burning coal definitely releases CO2
and other pollutants, including mercury and radioactive elements: in fact,
coal burning currently releases more radiation than nuclear power stations
(McBride et al., 1978) [5J].
A more relevant debate now is to compare nuclear power, which can
provide large quantities of continuous power from a single plant but which
may take decades to build, with ‘renewables’ such as wind and solar. Their
power production varies with the weather and requires many smaller in-
stallations to produce comparable power, but they can be built relatively
quickly.
An environmental debate for the mathematics classroom

Focusing on pollution and environmental issues rather than climate change,
Muddying the Waters [5K] is one of the more adventurous of the Mathemat-
ics Assessment Project’s Classroom Challenges. It is about modelling the pat-
tern of pollution changed by building a dam near a factory (Figure 5.5.1).
The lesson includes role-playing discussions between factory owners, the
owners of the Riverside Centre, prosecutors, and environmental officers and
critiquing the arguments and scientific evidence, quantitative and otherwise,
presented in a fictitious court case – an exercise in critical thinking.
Climate change 121
FIGURE 5.5.1 Extract from Muddying the Waters.

Source: Mathematics Assessment Project [5K].
5.6 Strategic issues and a critical thinking approach

This book is about how mathematics education can improve mathematical
literacy. How can it best contribute in the huge field of the human effects on
the environment, with its complex interacting phenomena? Trade-offs, and
therefore value-based decisions, are centrally involved. There are mismatches
of scale both in space and in time: local short-term needs versus global dec-
ade-by-decade changes. Established habits and political systems regularly
overmatch cumbersome international structures like the ‘Conferences of the
Parties’ (COPs), which require unanimity.
This strategic issue is a central concern of a community that takes a critical
thinking approach to the epistemological and ethical issues involved. Barwell
et al. (2022) point out that the human species is privileged over all others in
current thinking on the living world; instead, they commend a more symmetri-
cal ‘dialogic’ approach with ethics based on our answerability and responsibil-
ity in relation to all parts of the living world. This perspective stresses that all
parts of the ecosystem, including humans, are in a relationship with each other.
For those concerned with mathematical literacy in this context, mathematics
education is understood as including the conditions for learners to explore the liv-
ing world, with suitable guidance, and to become fluent in the appropriate forms
of enquiry. Ecosystem crises, including which actions seem needed, become situ-
ations to be explored and understood with an emphasis on participation.
This reflects the central message of this book: that mathematics educa-
tion must go beyond skills and knowledge of mathematics to prepare stu-
dents to be engaged citizens, able to engage with uncertainty, values-based
122 Climate change
decision-making, and democratic debate. Hauge and Barwell (2022) propose

three principles in support of such an approach: exploring meaningful situa-
tions of risk and uncertainty; exploring both scientific/mathematical concepts
and societal perspectives; and exploring and learning through dialogue. The
need for attention to risk and uncertainty and quantitative assessment of
change in this context is clear for the multiple ecosystem crises we face.
Acknowledgements
Our thanks to Richard Barwell for informative exchanges about the dialogic
perspective.
Figure 5.5.1 from the Mathematics Assessment Project is available under
the Creative Commons CC BY-NC-ND 3.0 licence and appears here courtesy
of the Bell Burkhardt Daro Shell Centre Trust.
Thanks to NASA, NOAA, and other agencies for the freely available
graphs and data used extensively in this chapter. Figure 5.2.1 is from Leven-
dis et al. (2020) published by the Royal Society under the Creative Commons
CC-BY-4.0 licence. All of these institutions offer reputable data on climate
change issues – any mistakes or misrepresentations in presenting this data
here are fault of the (non-climate scientist) authors of this book.
References
Arrhenius, S. (1896). XXXI. On the influence of carbonic acid in the air upon the
temperature of the ground. The London, Edinburgh, and Dublin Philosophical
Magazine and Journal of Science, 41(251), 237–276.
Barwell, R., Boylan, M., & Coles, A. (2022). Mathematics education and the living
world: A dialogic response to a global crisis. Journal of Mathematical Behavior, 68,
101013. https://doi.org/10.1016/j.jmathb.2022.101013
Hauge, K. H., & Barwell, R. (2022). Education for post-normal times. In R. Herheim,
T. Werler, & K. H. Hauge (Eds.), Lived democracy in education: Young citizens’
democratic lives in kindergarten, school and higher education (pp. 65–76). Routledge.
Levendis, Y. A., Kowalski, G., Lu, Y., & Baldassarre, G. (2020). A simple experi-
ment on global warming. Royal Society Open Science, 7(9), 192075. https://doi.
org/10.1098/rsos.192075 [5B]
MacFarling Meure, C., Etheridge, D., Trudinger, C., Steele, P., Langenfelds, R., van
Ommen, T., Smith, A., & Elkins, J. (2006). Law Dome CO2, CH4 and N2O ice
core records extended to 2000 years BP. Geophysical Research Letters, 33(14),
L14810. https://doi.org/10.1029/2006GL026152. Data set retrieved from https://
www.ncei.noaa.gov/access/paleo-search/study/9959
McBride, J. P., Moore, R. E., Witherspoon, J. P., & Blanco, R. E. (1978). Radiologi-
cal impact of airborne effluents of coal and nuclear plants: Radiation doses from
airborne effluents of a coal-fired plant may be greater than those from a nuclear
plant. Science, 202(4372), 1045–1050.
Sawyer, J. S. (1972). Man-made carbon dioxide and the “Greenhouse” effect. Nature,
239, 23–25.
Climate change 123
Supran, G., Rahmstorf, S. & Oreskes, N. (2023). Assessing ExxonMobil’s global

warming projections. Science 379(6628). https://doi.org/10.1126/science.abk0063.
Retrieved from https://www.science.org/doi/10.1126/science.abk0063 [5E]
Thom, R. (1977). Structural stability, catastrophe theory, and applied mathematics.
SIAM Review, 19(2), 189–201.

purposes:
[5A] Global Climate Change – NASA

https://climate.nasa.gov/vital-signs/global-temperature/
[5B] A Simple Experiment on Global Warming – Royal Society Open
Science
https://royalsocietypublishing.org/doi/10.1098/rsos.192075
[5C] Showing that Carbon Dioxide Is Made When a Hydrocarbon Is Burned
https://youtu.be/6l8D6DsRDYk
https://youtu.be/6o8QXvKZ2xI
https://youtu.be/b7e9QjNX51s
[5D] Global Warming Potential – Wikipedia
https://en.wikipedia.org/wiki/Global_warming_potential
[5E] Assessing ExxonMobil’s Global Warming Projections – Science
https://www.science.org/doi/10.1126/science.abk0063
[5F] Trends in Atmospheric Carbon Dioxide – NOAA Global Monitoring
Laboratory
https://gml.noaa.gov/ccgg/trends/
[5G] You Reckon? Bio Fool – Bowland Maths
https://www.bowlandmaths.org/materials/projects/online/you_
reckon/You%20Reckon_Web/page_17.htm
[5H] Energy Density of Uranium – various sources
https://www.plux.co.uk/energy-density-of-uranium/
https://atomicinsights.com/energy-density-comparison/
https://www.euronuclear.org/glossary/fuel-comparison/
[5J] Do Coal-fired Power Stations Produce Radioactive Waste? – Scientific
American
https://www.scientificamerican.com/article/coal-ash-is-more-
radioactive-than-nuclear-waste/
[5K] Muddying the Waters – Mathematics Assessment Project
6
PLANNING FOR GOOD THINGS IN LIFE
We all know, in principle, that careful planning of what we do with our time
can enable us to get more out of life. But the planning itself takes time – and
for many people it is an uncongenial activity. They would rather “be getting
on with life” without having to spend time planning the details of what to do
when, and how to make it happen. Yet some things obviously need to be
planned – or else frustration, or worse, is sure to follow. So what kinds of
planning are right for you?
ACTIVITY
Ask small groups to discuss what situations they plan for – and for what other
situations they think it might be worth it.
Lifestyle choices and opportunities are obviously determining factors in

the range of options considered – the more choices you have, the more deci-
sions you have to make as to what to do, and when. Some total ‘spur of the
moment’ people don’t plan, but if you have some time for ‘slow thinking’,
then considering what and how to plan is worthwhile.
All work environments involve planning of various kinds. For junior
employees their work may be largely imposed ‘from above’, though in any
well-run company there is space for all employees to use their intelligence in
making the management’s plan work as well as possible. Promotion leads
to increased responsibility for planning decisions – for example, in detailed
DOI: 10.4324/9781003303503-7
Planning for good things in life 125
scheduling and in assigning people to tasks in a way that reflects their differ-
ent capabilities.
In this chapter we first look in Section 6.1 at the challenges in organis-
ing events and how mathematics can be useful. Section 6.2 outlines some
well-engineered teaching materials focused on planning. Section 6.3 looks
at mathematics to tackle the more general problem of optimising a situation
within constraints, extending this in Section 6.4 to the science and mathemat-
ics of logistics – with a forward look to data science, to which we return in
Chapter 7. The chapter includes a mix of ‘know how’ and ‘know about’
mathematical literacy mixed with critical thinking. Planning is something
nearly everyone must know how to do, and it is also useful to know about
the sophisticated tools available for complex situations.
6.1 Organising events

The Airplane Turn-round and Table Tennis Tournament tasks [6A] intro-
duced in Chapters 1 and 2 are typical of a genre: organising events or, more
widely, processes. Other familiar examples are as diverse as getting the fam-
ily up and ready for school or work, or organising a school event, an election,
or a strike. The ability to organise is a valuable high-level skill. What are its
main elements?
• The participants
• The physical environment in which the event will take place
• The sequence of activities and the rules that govern them
The challenge is to devise a systematic approach which addresses all the

significant variables and, one hopes, all eventualities that might arise. Usu-
ally, the organiser will have knowledge of similar events – even experience in
taking part – but they are unlikely to have thought through all the aspects of
planning required.
It is easy, and usually disastrous, to forget to carefully consider the inter-
ests and feelings of the people who will be involved; the study of ‘human fac-
tors’ is a field of its own, beyond the scope of this book except to underline
the importance of trying to get everyone ‘onside’. In the same way, a congen-
ial physical environment is an important factor in the success of any event.
These matters are not mathematical beyond being included in a systematic
approach.
For the sequence of activities and the rules that govern them, simple math-
ematics is helpful. Things like lists and tables are essential. Less familiar types
of diagrams can clarify the situation. For example, the Airplane Turn-round
task (introduced in Section 1.3) lists seven operations taking a total of
160 minutes (Table 6.1.1) and asks how this time can be minimised. In this
126 Planning for good things in life
TABLE 6.1.1
Operations needed to ‘turn-round’
a passenger aircraft
Operation Time
(minutes)
A. Disembark passengers 10
B. Clean the cabin 20
C. Refuel the plane 40
D. Unload the baggage 25
E. Embark new passengers 25
F. Load new baggage 35
G. Final safety check 5
Source: Adapted from Bowland Maths [6A].
case, with the realisation that some of the jobs can be performed in parallel,
one can sort out the minimum time needed by some thought and some trial
and error.
ACTIVITY
Ask small groups to tackle the Airplane Turn-round task individually then dis-
cuss how they did it.
Systematic representations – Pert diagrams and Gantt charts

There are various systematic methods for tackling scheduling problems
of this kind. The ‘Pert diagram’ in Figure 6.1.2a shows one systematic
way to record which steps must be sequential and which can be done in
parallel, and the times involved. There is a formal Program Evaluation
and Review Technique (Pert) for using such diagrams in more complex
situations.
The total turn-round time is determined by the path through the dia-
gram with the longest total time – in this case 65 minutes for the steps of
unloading and loading baggage and the final check. So, efforts to shorten
turn-round time should focus on the steps on this critical path. Knowing
to identify the critical path, and try to shorten it, is a powerful general
insight – and strategy.
If, for safety reasons, it was decided that refuelling should not take place
while any passengers are on board (as with rockets) the modified diagram
Figure 6.1.2b would apply. The critical path changes so the turn-round
time becomes 80 minutes – an additional operating cost for the airline.
FIGURE 6.1.2 Simplified Pert diagrams for refuelling (a) with and (b) without pas-
sengers. ‘Critical paths’ shown in bold.
Cooking and serving a large meal with many dishes is a situation of similar
complexity that many people will have grappled with.
Another planning tool in regular use is a Gantt chart. Gantt charts arrange
the tasks along a timeline, which makes them useful in organising the details
of complex projects and monitoring their progress. Figure 6.1.3 provides a
simple example, based on Figure 6.1.2b. Gantt charts can be constructed in
spreadsheets, though specific products with extra functionality are available.
These examples again show the value of customised mathematical represen-
tations in giving power over complex situations.
FIGURE 6.1.3 Airplane turn-round – a Gantt chart.

ACTIVITY
Plan how to cook and serve a meal on time.

Choose recipes for a three-course meal (appetiser, main course, and des-
sert) and draw a suitable Pert diagram or Gantt chart. Identify the critical path
and when you should start preparing for a 7 p.m. dinner. Do not include prep-
aration for components (e.g. a cake) that are ready in advance. It will save
time to start by writing the steps on individual ‘sticky notes’ (or the computer
equivalent) to work out the critical path before trying to draw a neat diagram.
Pedagogical challenges
The following are some of the issues teachers face in developing these skills
for organising events:
• How open should the task be?

• How much time should be devoted to the activity?
• How should support be given to those who struggle with the task, or when
their work no longer seems productive?
• What is the payoff in learning?
These are much the same challenges as arise when solving non-routine
problems within mathematics. Let us look at them in the context of organis-
ing a tournament. Teachers may feel that the version of Table Tennis Tour-
nament in Chapter 2 (Figure 2.1.2) – which simply asks students to plan a
tournament – is too open, though it will often arise in this vague form in
real-world situations, leaving the organiser to make it more specific. A lightly
scaffolded version designed for use in a single lesson is shown in Figure 6.1.4.
Notice the questions that an organiser should ask in embarking on this
task are not only raised but answered in this version. This reduces the stra-
tegic demand of the task, assists discussion by setting common parameters,
and gives the teacher a more predictable lesson, and certainly makes for an
easier job when checking students’ solutions! The version in Figure 6.1.5
takes this further.
In Figure 6.1.5, Q1 may well be just an exercise, testing skills that have
been taught. Q2 structures a solution, setting out the parameters and show-
ing exactly what a solution includes. It can be tackled by trial and error – and
usually is – but there is an elegant systematic, generalisable solution based on
keeping one player fixed and cycling the others.
The Bowland Maths professional development module on Tackling un-
structured problems in the classroom [6A] begins with an introductory ses-
sion in which teachers discuss the issues, compare more or less structured
You have the job of organising a table tennis league.

• 7 players will take part.
• All matches are singles.
• Every player has to play each of the other players once.
• There are four tables at the club.
• Games will take up to half an hour.
• The first match will start at 1.00 p.m.
Plan how to organise the league, so that the tournament will take the shortest
possible time.
Put all the information on a poster so that the players can easily understand
what to do.
FIGURE 6.1.4 Table tennis tournament – a lightly scaffolded version.

Source: From Bowland Maths: Tackling Unstructured Problems [6A].
FIGURE 6.1.5 Table tennis tournament – a directive version.

Source: From Bowland Maths: Tackling Unstructured Problems [6A].
versions of the task, watch a lesson video, and then together prepare a lesson
based on a relatively open version of the task to teach. After observing how
the activity works, they reflect in a structured follow-up session on what hap-
pened and what they have learned about handling non-routine problems in
the classroom.
6.2 A case study for teaching planning and organising

This section gives an example of materials to develop planning skills. The
Numeracy through Problem Solving project (NTPS) was described in
Section 2.4. Three of the five modules, Design a Board Game, Produce a
Quiz Show and Plan a Trip, target planning skills. For those who want a
well-supported entry into teaching for mathematical literacy, these modules,
and their four-stage design, may be useful.
Each module works on a group-project basis and has four stages:
• Stage 1: Explore the domain by working on and evaluating exemplars

provided.
• Stage 2: Generating and sifting ideas, then reviewing them and deciding
what to take forward.
• Stage 3: Developing a ‘product’ in detail.
• Stage 4: Evaluating what other groups have produced – then reviewing
your own.
Produce a quiz show

Many websites provide support for game show simulations of various kinds.
In the NTPS module, student groups devise, schedule, run, and evaluate their
own game shows. This involves: preparing, timing, and testing questions (us-
ing number and statistical concepts); planning room layouts; and scoring
systems. There are four stages:
• Stage 1: Groups of students take it in turns to act out a number of TV-

type quizzes that are provided, identifying and commenting on faults and
shortcomings in the organisation, rules, questions, scoring systems, and
presentation. These games are designed with faults – some obvious, others
less so; using established game shows as a model would set a standard that
can be discouraging.
• Stage 2: In groups, students share ideas for their own quiz, reach agree-
ment on which to develop, and draw up a plan of action.
• Stage 3: Each group prepares, tests, and organises its questions, scoring
systems, rules, and final running order. Groups decide how the furniture
and equipment will be arranged during the show.
• Stage 4: Groups take it in turns to present their quizzes, with the rest
of the class providing competitors – and the audience. Afterwards, each
quiz is evaluated, first by other members of the class, and then by the
group who produced it. A further opportunity may be given for a group
to enact their quiz with different groups of contestants – perhaps a dif-
ferent class.
The student booklets are designed to maximise student agency, providing

suggestions on essentials only after the point where students should have
thought of it for themselves. All the Numeracy through Problem Solving
materials can be downloaded free from the Shell Centre website [6B].
6.3 Optimising within constraints

In planning, there are always constraints of time, money, personnel, and
available resources – just to name the most obvious. Often the constraints
sharply limit the range of options available, so seeing how far the constraints
can be stretched is a strategy for getting a better outcome.
Vacation planning is a common example, where most people have con-
straints, including those on both time and money. Your job probably comes
with a fixed number of paid vacation days off work each year; if you can
take your vacation in weeks that contain a public holiday, you can be away
from home for longer. Equally starting out on a Friday evening and returning
just in time for work on a Monday morning will extend your vacation. But
paying for the extra accommodation may challenge your money constraint.
In this way, different constraints compete for priority, which leads to the
question of what you are seeking to optimise. Suppose you are looking for
reliable sunshine, warmth, and a beach. That will eliminate a lot of nice
places and help to rank order those that satisfy this constraint. When you
consider the cost, and also want to be in an interesting place, other trade-offs
come in. How are these things (‘goods’ to economists) to be balanced?
Most of us who try to be somewhat systematic would, when planning such
a vacation,
• explore the range of possibilities, for example by browsing the web and
talking to friends;
• choose some ‘candidates’ that fit within our ‘tight’ constraints;
• gather information on cost, weather, and attractions; and then
• make a choice based on ‘an overall impression’.
This method seems to satisfy people – the main aim. Discontent usually
arises when the data are wrong: the flight is delayed, the hotel is still under
construction, or the “nearby beach” on the website is a mile away.
Systematic approaches
For situations where more formal, well-defined, methods of choosing are
needed, there are various approaches based on two fundamental concepts:
boundaries and utility. The concept of boundaries is a relatively straightfor-
ward one – tight constraints that must be satisfied by any solution. The PISA
item Which car? in Figure 2.2.1 is an example where the boundaries alone
determine the choice. More often there are many possible choices that fit the
boundaries; then the concept of a utility function comes in. This is a function
of all the significant variables that reflect the values of the person or other
entity making the decision. For some situations, like those focused on money
which we will address in Chapter 9, the utility function may be relatively
straightforward. The directors of a British or American public company are
currently required by law to focus predominantly on ‘shareholder value’ –
so the share price is the utility function. Some jurisdictions also include the
interests of other ‘stakeholders’, both employees and society at large, so de-
ciding on the balance between competing ‘goods’ is part of the management
challenge.
In such situations, a function must be built from the significant variables.
Unsurprisingly, a linear function is often preferred, so the utility U is defined by
U = w1v1 + w2v2 + w3v3 +  + wn vn
where wi is the (numerical) weight assigned to the variable vi .
ACTIVITY
Ask small groups to apply boundaries and utility to planning a short vacation:
• Set three boundary constraints for the vacation (e.g. maximum cost per
person, dates, main activity) and choose at least three utility variables (e.g.
travel time, daily temperature, live music venues).
• Find five alternative vacations that fit the boundary constraints (alternatives
might be variations, e.g. in dates).
• Compare the alternative vacations using the informal approach described
above and choose one.
• Construct a set of weights for each variable that will make the preferred
choice have the highest utility. (Hint: In deciding weights it is the variation
in each variable that determines its overall influence on U, not its absolute
value).
• Discuss the experience.
Phil and Cath plan to make and sell boomerangs in two sizes: small and large.
Phil will carve them from wood, taking 2 hours for a small one and 3 for a large one.
Phil has a total of 24 hours available for carving.
Cath will decorate them.
She only has time to decorate 10 boomerangs of either size.
The small boomerang will make $8 profit, the large boomerang $10.
They want to make as much money as they can.
How many small and large boomerangs should they make?
How much money will they then make?
FIGURE 6.3.1 Maximising profits: selling boomerangs.

Source: Mathematics Assessment Project [6C].
This domain of mathematics, linear programming, is a standard topic in

some high school curricula. Perhaps surprisingly, the core result is that the
boundaries usually play a central role in determining the optimum.
Boomerangs – an optimisation activity

The lesson Maximising profits: Selling boomerangs illustrates this kind of plan-
ning. The essence of the problem is in Figure 6.3.1. The goal is maximising
profits (the utility function) within two boundary constraints of time available.
The materials on the MAP website [6C] provide detailed support, with ex-
amples of student work for discussion in class. It assumes that students have
not yet been taught linear programming, so this is a non-routine problem for
them to solve. Most students tackle this novel problem by enumerating pos-
sibilities and then organising them in a table – a more challenging task than
one might think. The four examples of student work in the module, and the
commentary on them, are illuminating.
A more sophisticated approach using algebra, with x small and y large
boomerangs, expresses the profit and constraints as
• P = 8x + 10y (profit)
• x + y ≤ 10 (Cath’s time constraint)
• 2x + 3y ≤ 24 (Phil’s time constraint)
• x ≥ 0 and y ≥ 0 (no negative boomerangs!)
When there are only two variables, the standard linear programming
approach is to graph these linear functions, looking for the maximum value
of P which stays with the constraints. Figure 6.3.2 shows how this works –
any values of x and y falling in the white area break one or both of the time
constraints. The ‘permitted’ area is the darkest shaded region ABCD. P = 88
FIGURE 6.3.2 Maximising profit on ‘boomerangs’ – a linear programming solution.
is the maximum value for which the ‘profit’ line P = 8x + 10y passes through
ABCD, just ‘kissing’ it at point C where there are six small and four large
boomerangs.
A graph like that in Figure 6.3.2 can be easily made with free online tools
such as Desmos Graphing Calculator or GeoGebra [6D]. The website links
to one. First plot the two inequalities, which can be typed in as ‘x + y ≤ 10’
and ‘2x + 3y ≤ 24’. Then, add the profit formula as ‘8x + 10y = P’. Desmos
and GeoGebra will automatically create a ‘slider’ that lets you interactively
adjust P, although you will need to specify the slider settings to allow a suf-
ficiently large range for P.
As an extension of this problem, it is interesting to vary the parameters
(10, 24, 8, and 10) to see when the optimum moves to a boundary – either
all small or all large boomerangs, so at a different corner of the outlined
shape – or when there is more than one optimum value of x and y. It is easy
to explore these questions using the graphing calculator sliders.
6.4 Logistics and data science

Linear programming is just one approach in a broad field of mathematical
approaches to doing things more efficiently. This work goes back to the eight-
eenth century, but it really escalated during World War II under the name of
Operational Research (OR, also called Operations Research). In 1947 under

the auspices of the British Association for the Advancement of Science, a sym-
posium was organised in Dundee. In his opening address, Sir Robert Watson-
Watt offered a definition of the aims of OR (Zuckerman, 1964):
[T]o examine quantitatively whether the user organization is getting from

the operation of its equipment the best attainable contribution to its over-
all objective.
(p. 288)
This down-to-earth definition reflected his role, as one of the inventors of

radar, in getting as much useful information as possible out of the limited
numbers and capabilities of early “radiolocation” sets. The most famous case
was in the Battle of Britain where radar played a major role in the efficient
deployment of defensive fighter aircraft, which in turn prevented an air inva-
sion of Britain. If you generalise ‘equipment’ to ‘resources’, this remains a
fair definition of the field now usually called logistics. A great variety of ap-
proaches have been developed since then. They involve collecting data, then
mixing modelling and data analysis in various ways.
Making sense of data has a long history, often carried out by scientists,
statisticians, librarians, computer scientists, and many other professions, as
well as mathematically literate citizens in their everyday lives. To take one
foundational example, in 1854 there was a cholera epidemic in London.
John Snow (1854) collected data and made a ‘map’ of the address in London
where each case occurred – part of his map is shown in Figure 6.4.1. Each bar
on the map represents one death at that address. The cluster near the Broad
Street water pump alerted him to the cause of the cholera outbreak – the
FIGURE 6.4.1 Part of John Snow’s map of the 1854 London cholera outbreak.
Each bar represents one death at that address [6E].
water supply, rather than “unhealthy air from the river” which was previ-
ously thought to be the cause.
This work was the beginning of systematic epidemiology. Snow’s vivid dis-
play informed how we visualise and analyse data to this day. In the twenty-
first century, advances in computer power, the digitisation of the world
through the internet, and associated software design (including artificial
intelligence) enable data scientists to extract highly complex patterns from
enormous quantities of data and so develop probabilistic predictive models
that allow better business decisions. But the underlying principles are the
same as those applied by John Snow in 1854.
The principles of data science

We illustrate the principles of data science with a very simplified example.
Let’s suppose we run a small business – an ice cream shop. We will use a data-
driven logistics procedure, also called predictive modelling.
Step 1 is to identify your system of interest – a system being just some ‘bit’ of
the world, in our case the ice cream shop with its surroundings, and its possible
customer base of people who live near or travel by. We also need to identify
which variables – also called properties, or attributes – are important for our sys-
tem. For the ice cream shop, the number of ice creams sold would be of central
importance. Other variables, such as the weather, are also likely to be important.
Step 2 is to measure the system, to collect data. This could be as simple
as recording the number of ice cream sales each day, along with the daily
temperature, including the forecast for Friday. This data is stored in a spread-
sheet on a computer, or in a database, as in Figure 6.4.2a.
Step 3 is to find a mathematical function which takes in the things we
know and predicts things we want to know. We know the ice cream sales and
temperature each day in the past – and we know temperature forecasts for
tomorrow (Friday) are reasonably accurate. We want to predict the number
of ice creams we’ll sell tomorrow as indicated in Figure 6.4.2b. From experi-
ence, we might suppose that the hotter the weather, the more ice cream we
sell. We then look for a function (or model – here they’re the same thing)
that expresses the relationship between temperature and ice cream sales.
FIGURE 6.4.2 The essentials of data science.

FIGURE 6.4.3 A predictive model for ice cream sales.
We are free to find this mathematical function by any means – experiment

with functions using pen and paper or a spreadsheet, a statistician might
make statistical models, and a data scientist might use machine learning.
From these four data points, it is clearly not a linear function – indeed, it
looks as though the sales roughly double for every 5°C increase which would
predict 82 sales of ice cream on Friday, shown numerically and graphically
in Figure 6.4.3.
From a data science perspective, how we find this function is not funda-
mentally important; what matters is whether the function makes sufficiently
accurate predictions. We can test a candidate predictive function by using
it to “predict” our historical data, and then see how closely its predictions
match our records of the past. As in all data-driven research, it is important
to generate the hypothetical model with one data set and then test it on an-
other one. In this case, we might test the model on data from a friend’s ice
cream shop. This key principle is often ignored. (Of course, this four-point
data set is ridiculously small and only used to illustrate the principles.)
Step 4 operates when we find a predictive function that predicts with suffi-
cient accuracy. Then the final step of the process is to use this model to predict
the future and make better informed decisions. In our case, we could predict
how many ice creams we will sell on Friday. This could be of great value to our
decision-making; we could check to see if we have enough ice cream in stock
or need to order more, we might roster more staff to work today, or perhaps
on other days not bother opening at all because we don’t expect enough sales!
A data scientist might extend this model to include many more variables –
the location of the ice cream store, nearby events such as a local park fair,
which flavours sell best, or online marketing activity, with the hope that this
would give a function with better predictive power. The difference between
humans making estimates in their head, researchers in Operational Research,
scientists forming theories of the world, and machine learning practitioners
is only a difference in the tools they use and the scale of the enterprise. The
objectives and procedural frameworks are the same. Modern data science,
appropriately called ‘big data’, enables these principles to be extended to
extremely large numbers of variables and data points.
Acknowledgements
We are grateful to Daniel Burkhardt Cerigo for contributions to our thinking
about planning and data science, including Figures 6.4.2 and 6.4.3.
Table 6.1.1 and Figures 6.1.4 and 6.1.5 are used with the permission of the
Bowland Maths maintainers. © Bowland Charitable Trust 2008. Bowland
Maths materials are free for educational use.
Figure 6.3.1 and extracts from Selling Boomerangs from the Math Assess-
ment Project Classroom Challenges [6C], courtesy of Bell Burkhardt Daro
Shell Centre Trust. The full materials are available under the Creative Com-
mons Attribution, Non-commercial, No Derivatives Licence 3.0.
Figure 6.4.1 is taken from a version of Snow’s (1854) map digitised by the
UCLA Department of Epidemiology – see link [6E] for the full version.
References
Snow, J. (1854). On the mode of communication of cholera. C.F. Cheffins, Lith.
Zuckerman, S. (1964). In the beginning – and later. OR, 15(4), 287–292. https://doi.
org/10.2307/3007115.

purposes:
[6A] Tasks from Professional Development Modules – Bowland Maths

Table Tennis Tournament:
https://www.bowlandmaths.org/materials/pd/online/pd_01/pd_01_
class.html
Aircraft Turn-round: https://www.bowlandmaths.org.uk/materials/
pd/online/pd_05/pd_05_class.html
[6B] Produce a Quiz Show – Shell Centre for Mathematical Education

https://www.mathshell.com/materials.php?&series=numeracy
[6C] Maximising Profits: Selling Boomerangs – Mathematics Assessment
Project
[6D] Graphing Tools for Linear Programming – Desmos & GeoGebra
https://www.desmos.com/calculator
https://www.geogebra.org/calculator
[6E] Map of Cholera Cases in London – Snow (1854)
https://www.ph.ucla.edu/epi/snow/highressnowmap.html
https://https://en.wikipedia.org/wiki/John_Snow
7
LOOKING PAST THE ‘SPIN’
Every day we are bombarded with messages from people who would like to
sell us something – everything from products we don’t need but might enjoy
to political ideas that might persuade us to support their party or their cause.
They often set out to make their case more persuasive by quoting numbers or
displaying graphs that support it. But these are, inevitably, carefully selected
from a very much larger set of data. How does a mathematically literate
person best think through the options implied, to ‘buy’ or not to buy – as
ever, without wasting time? That is the theme of this chapter. Much of what
follows will be familiar to the reader, but we hope a broad overview will be
thought-provoking.
Context is always an important part of mathematical literacy, but
particularly so here – where mathematics is used to add credibility to an
argument. ‘Looking past the spin’ involves critical thinking that takes the
context seriously – looking in depth at the context, the language used to
present the argument, and how mathematics is being used to support it. Is
the mathematics correct in itself? Are the assumptions behind the mathemat-
ics, such as the data chosen, valid? Does the language used fairly describe
what the mathematics shows? Consequently, this chapter spends significant
time discussing the wider background of topics where debate is frequently
supported by the use, or abuse, of mathematical data.
With an exception that we will return to later, the people who design
marketing – whether in business or politics – usually avoid ‘the lie direct’. To
present as a fact something that can be shown to be untrue runs two risks:
rebuttal by a trusted ‘fact checker’ and, much more seriously, damage to your
reputation as a reliable source. That is a long-term injury that few organisa-
tions think worth risking for a short-term gain in sales or influence.
DOI: 10.4324/9781003303503-8
Looking past the ‘spin’ 141
7.1 Comparing products

Let us start with the most obvious form of marketing. Advertisements for
‘consumer products’ assail us on the street, in every corner of the internet,
and of our television and screen viewing. Even non-commercial stations like
the BBC advertise their own programmes extensively.
From the marketers’ point of view, this is a public service, letting us
know about things we might well have missed. This is undoubtedly true and
potentially valuable. But from the public’s point of view, it is not something
we can avoid or easily control. It sometimes feels, in a memorable phrase,
‘like cocktails from a firehose’. Friends add to this with accounts of things
they have bought or looked at. How does the mathematically literate person
handle this stream of information – and disinformation?
We examine in Chapter 9 the issue of ‘needs’ versus ‘wants’. Here we just
assume that you are well aware of your needs. If an advertisement should
remind you of a need that you have forgotten ‘to put on the list’, so much the
better – a real public service, epitomised in ‘public service announcements’
designed to warn of risks or encourage healthier living. But, beyond needs,
what about things you would simply like to own or do?
Most people have a ‘small’ sum of money that they are prepared to spend
without worrying about it. For things under that limit, it is time-effective just
to buy it. Of course, you will review your limit from time to time – particu-
larly if your circumstances change. You will also have to consider how ‘small’
outlays accumulate over time – there are marketing practices that exploit this
by advertising a modest-sounding subscription, or payment in instalments.
If you have created a weekly or monthly budget for ‘discretionary spend-
ing’ – on clothes, tickets, or whatever – some agonising may be in order.
Indeed, it may be part of the enjoyment, perhaps with friends, to talk through
various things you might buy. But again, many would feel careful considera-
tion of alternatives would spoil the fun. Personal budgeting of this kind is
discussed in more detail in Chapter 9.
For more substantial purchases, finding the right product at the right
price demands more thought. The potential saving from ‘shopping around’
for a new smartphone or television could fund another substantial ‘want’.
But looking at alternatives, evaluating their quality and comparing prices is
potentially a huge task. Fortunately, there is help at hand. The traditional
sources are consumer magazines and their websites – sites such as Con-
sumer Reports in the United States, Which? in the United Kingdom, Choice
in Australia; these have established programmes of regular research-based
comparative reviews of consumer items from coffee grinders to cars – but,
typically, you will need a paid subscription to get beyond the free overview.
These have been joined by a large number of product comparison websites
offering professionally written reviews – although, since these are typically
142 Looking past the ‘spin’
funded by advertisements, possibly for the very products they are reviewing,
users need to assess their impartiality.
This ‘traditional’ approach to product reviewing now faces its own com-
petition. With the increase in online shopping, ‘crowd-sourced’ user reviews
and ratings have become ubiquitous – and being able to hear the honest
views of genuine people who have actually bought and used the product
sounds wonderful – but some are fake reviews from ‘users’ who are given
incentives to write positive reviews, or negatively review a competitor (see,
for example, The Secret World of Fake Online Reviews, The Guardian, 22
April 2023 [7A]). Even without deliberate fakery, there are sources of bias –
a proportion of all goods are faulty and someone angry at receiving a faulty
unit may be more inclined to write a review than many users who thought the
product was OK. Others may rush out a glowing review based on first impres-
sions that misses a serious flaw in the product. “Star” ratings are popular –
but what do they mean? Some people regard anything less than five stars as
a failure, others will see anything more than three stars as acceptable and
reserve five stars for exceptional products – and if a product has hundreds of
reviews, it is tempting to go by the average rating. Table 7.1.1 parodies the
problem of superficial five-star reviews.
More recently, social media have become a leading source of product
information for consumers – and, hence, a vital marketing tool for business
and politics. Initially the domain of enthusiastic amateurs, being a social media
‘influencer’ can now be a lucrative full-time business with celebrity status, uti-
lising professional videos with high production values. A good influencer will
be totally transparent about sponsorship and how they obtained the products
they feature. They often produce in-depth reviews rivalling the traditional
consumer advice publications. Others will give the impression that these are
products that they use personally (true or false) which they are recommending
to you as a friend.
TABLE 7.1.1 How do you summarise reviews?
Average customer review:

Nev Loved the colour
Kat87 Arrived next day, well packaged
Wurzel Fits like a glove, came with free pen
Joolz Does what it says on the tin
Fluff77 Looks super, but batteries not included in box
JWSmith Exploded – second-degree burns. House burned

down. Posting this from hospital
For illustrative purposes only – as they say in the ads.
The internet has brought another, often neglected, option, especially when
buying ‘tech’ products or appliances: it is now often possible to download the
instruction manual, which often gives clearer, more detailed information about
the product’s specifications and capabilities than the marketing material.
How can you make objective decisions?

The wide range of products and the variety of features that are so often on
offer can make thinking through alternatives in a systematic way a challenge.
The mathematical challenge of integrating the quality assessments from
reliable sources with the prices on offer is modest – basically one of organis-
ing information. ‘Irresistible offers’ at the ‘sales’ complicate the process only
slightly.
Whereas systematic reviews are objectively worthwhile, it is not surprising
that many of us just look for a good discount from a source we know and
have found reliable, and buy.
ACTIVITY
How do you make buying decisions?

Students compare notes on how they make buying decisions at various cost
levels.
In Chapter 6 we looked at a systematic method of comparing things with

various desirable attributes by adding individual scores ( vn ) with weights ( wn )
chosen to reflect the importance of that factor – and how much those scores
vary – defining a ‘utility’ function:
U = w1v1 + w2v2 + w3v3 + …+ wn vn
The same approach is used in evaluating products. Looking back at the

simplistic averaging in Table 7.1.1, you might feel that product safety deserves
more weight than colour or packaging. To take a real example, an article from
Choice [7B] describes how they rank washing machines. For each m achine,
scores are allocated to performance on five variables: dirt removal (d), rinse
performance (r), gentleness (g), water efficiency (w), and spin efficiency (s).
Each of these variables would have a separate scoring system. Then machines
are compared using the utility function:
U = 0.4d + 0.2r + 0.15g + 0.15w + 0.1s

The system to decide the winners of the decathlon in track and field athlet-
ics is a complex real-world example of a utility function – see this article from
NRICH [7C]. There are ten events. Each athlete receives a score for perfor-
mance (either time or distance) on each event. The scoring systems, defined
and revised separately for each event, were originally set up so that a world
record performance would receive 1 000 points. For the decathlon, events are
equally weighted, so the scores are added to determine the winner. Other
sports, like fantasy or real football, offer many examples.
ACTIVITY
Creating and using a utility function.
a How would you define the GOAT (greatest of all time) for the sport, or the
music style, of your choice.
First, in groups choose a sport, music style, or some another shared
interest. Decide on the contenders and the variables (factors) which make
‘greatness’. Decide how each factor should be scored and assign weights to
each factor. Then create a U-value for each individual. Discuss whether you
agree with the rank order your U-function gives. If not, try changing the
weights so that it better represents your values.
b Alternatively, undertake the same activity for products. Choose a ‘product
category’ (e.g. coffee machines, electric luxury cars) for which there are
several factors which significantly affect desirability, and on which there
is easily accessible data for multiple individual products. For example, for
electric luxury cars, factors will include range, style, internal fittings, and
price. Proceed as above.
In some countries test scores are used in this way to evaluate teachers
and schools. A narrow selection of variables leads to a distortion of what is
taught to focus on the content being tested, which only rarely includes all the
learning goals in a balanced way – ‘what you test is what you get’.
What do marketers do?

It is informative to think about marketing from ‘the other side’. The essence of
marketing has two elements – developing products that customers want and
persuading people to buy them. The ‘people’ here come in layers, for there
is usually a merchant of some kind between the manufacturer and the ulti-
mate customer and often a ‘supply chain’ of successive buyers: international
buyers who buy tea, say, from the growers, shippers who arrange transport
to wholesale markets around the world, wholesale buyers who blend and
package it for sale to shops, which sell it to us. (As a result, the tea grower
gets only a small fraction of the ultimate price; those who tend and pick the
tea much less.)
At each stage there is a market operating in which the sellers at that stage
compete with each other on quality and price. Persuasion backed by data
is at the heart of this: convincing the buyer that ‘our product’ will make
you more money than ‘theirs’. Manufacturers spend money on gathering
data to support their case – though not nearly as much as on advertising it!
These principles apply to products of all kinds, from cabbages to prescrip-
tion drugs.
Price setting
Marketers decide on a selling price in the context of ‘the 4 Ps’ – product,
price, place, and promotion – alongside an understanding of the market and
the ‘positioning’ of the brand within it. For basic products, like those sold in
supermarkets, ‘market and brand analysis’ takes into account many variables:
• The current portfolio of our products and prices.

• Where this product might best fit in the market, where we price it, and
why.
• Comparing it to key competitors’ similar products.
• Looking for a unique benefit or added value to warrant a premium.
• Consumer research – what consumers are willing to pay, any ‘price ceil-
ing’ or not.
• Where it will be sold, and if there might be a different pricing approach
by channel – for supermarkets, convenience stores, or wholesale channels,
for example.
The actual costs in producing, delivering, and advertising the goods are, of
course, a significant factor – as is the company’s expectation of profit margin.
But the impact of a product’s contribution to the total brand profitability is
a key consideration. Even if the product does not make money, a company
may launch it as playing an essential role for the brand – perhaps to drive
volume and operational efficiencies, to satisfy a loyal consumer base, or to
protect market share. Some products may be sold below cost (or at least at
a reduced profit margin) as ‘loss leaders’ to attract customers to the brand.
Others, especially ‘extras’ sold after the customer has chosen the main prod-
uct, may be sold at a huge markup to compensate for a low initial cost. In
summary, marketers take a broad view based on the total brand profit and
loss. Marketing clearly involves a diverse range of modelling competencies
that characterise mathematical literacy in linking human understanding with

analytical and mathematical skills.
For premium and luxury products, consumers buying decisions are not
primarily based on price. The marketing approach has the same elements but
brand strength, its role and positioning in the market are key factors. If you
think of, say, Apple, consumers want to own these products because they buy
into the ‘Apple vision’ and design ethos, and/or have prior experience. Price
is probably low down in the ‘consumer purchase decision tree’. Apple can use
this in their pricing models through mark ups on their products – how big we
can only speculate but the profitability of the company gives a clue.
Similar principles apply to services, though for these the supply chain is
shorter, often only one link – as with your hairdresser or lawyer, for example.
But medical services, for example, can have quite complex chains – from
primary care physician (or ambulance) to a hospital with its diverse range of
expertise deployed in variously structured ways.
A consumer decision
The following activities reflect consumer decisions that will be familiar to
students. They will also serve to stimulate further reflection after the later
sections of this chapter.
Choosing a mobile phone is something of a cliché here, but smartphone
ownership is ubiquitous around the world, and making a good choice is com-
plicated by the way phones are sold, often bundled with finance plans and
service subscriptions. When you have chosen the actual phone and level of
service you want, even a single vendor will typically offer multiple plans for
the same phone. Then you may have a choice of vendors, or even the option
of buying the hardware and service from separate suppliers. Each service
provider usually offers a range of “tariffs” offering different permutations of
call options, data limits, and contract lengths. The concepts of making quan-
titative estimates of your actual needs, interpreting the marketing, working
out the total cost of ownership, choosing valid measures for making compari-
sons, and interpreting the results make this a powerful example of informed
decision-making. In some countries, where you live might force you to choose
a particular network, restricting the choice somewhat, but in other areas (such
as the UK and EU) there is a fairly competitive mobile phone market that gives
most people a choice of several networks and service providers offering options
at very different prices.
For these activities we’ll assume that the hardware marketing folk have
done their job well and everybody wants the amazing nuPhone 42b more
than their next hot meal (actually choosing between phones is a whole other
adventure, and any discussion of that will be out of date before this book is
printed). So, the question is, how do you pay for it and what service subscrip-
tion do you want?
ACTIVITY
Students discuss and research what level of phone service they need.
How many minutes of phone calls do you make in a month? Are any of
them abroad? How can you estimate this? Is making traditional phone calls
and sending texts even important to you – if not, how many gigabytes of data
do you need to download in a month? If you already have a phone, then you
probably have information on past usage.
Discuss the following example. A phone plan offers 125 GB of data per
month at speeds of up to 100 Mbps. (Remember to think carefully about units!)
• How many hours per month could you spend downloading at top speed?
Is this realistic?
• How many hours of music could you download per month?
• How many hours of video could you download per month?
Use the internet to look at some phone service plans (SIM-only contracts)
and see what choices there are.
It is probably simplest to concentrate on 12–24-month service-only con-

tracts at this stage to ensure like-for-like comparisons. If you choose “pay as
you go” schemes that typically have a time limit, often 30 days, you could
consider including monthly top-ups to have the equivalent of a contract. The
issue of contracts versus pay-as-you-go and the implications of making a
contractual commitment can be reintroduced after the second activity, which
introduces the total cost of ownership.
ACTIVITY
Students research the cost of a phone over several years.

Now you have an idea what sort of phone plan you need, how much will it
cost, in the long term, to pay for the phone and service? Collect data from the
internet and build a spreadsheet to record it in an organised way. Questions
you might want to ask are as follows:
• How much do you have to pay up front, and how much per month?
• How long are you locked-in to the contract? What is the total you will pay
in this time? Are you likely to keep the phone for longer?
• Can you save any money by paying more up front? How does this depend
on how long you keep the phone?
• Can you buy the phone more cheaply elsewhere and just get the service
from the phone company?
FIGURE 7.1.2 Spreadsheet to compare total cost of phone ownership.

Source: Spreadsheet available online [7D].
There is unlikely to be a single clear “right” answer – you will need to think
about how long you want to commit yourself for, how much you can afford to
pay up front. The important thing is to be aware of the ‘total cost of ownership’
(how much you will spend on the phone over its lifetime) in various scenarios.
Figure 7.1.2 shows an example spreadsheet, but it is important to look up cur-
rent data and possibly adapt the format of the spreadsheet, as the way phones
are marketed varies with time and between countries. In recent years, some
phone companies have made it clearer how much you are paying for the phone
and how much for the service – bundling them together (as shown in the exam-
ple) was common in the past, when commonly the phone company continued
to keep charging the same monthly rate long after the contract period had ended
and the cost of the phone has been paid off. Sometimes it is more expensive to
get the phone from an electronics store than a phone company – who make their
money back on the service charges – but that could still work out cheaper if you
keep the phone for longer and can shop around for a cheaper service contract.
7.2 Advertising
Many Western countries have laws that limit the exaggeration that is permitted
in advertising. In the UK, the Advertising Standards Agency has codes for various
areas of advertising (print and broadcasting in particular). At the top level, it states:
The central principle for all marketing communications is that they should
be legal, decent, honest and truthful. All marketing communications
should be prepared with a sense of responsibility to consumers and society
and should reflect the spirit, not merely the letter, of the Code.
There is a lot of detailed guidance that marketers have to follow, more or

less, as to what this means in different areas of advertising. Not surprisingly,
many advertisers test the boundaries of the regulations – for which some are
fined or simply forced to withdraw the ad.
The internet still retains something of the Wild West on advertising, as for
so much else. The central principle set out in the UK ASA Code and its interna-
tional equivalents remains the goal, but the challenge of turning it into workable
regulations remains elusive – and, to some, undesirable. Most of the focus of
regulation was initially on avoiding harm, particularly to children, and on pro-
tecting an individual’s data from commercial exploitation without their consent.
In the United States, the guarantee of freedom of speech – a precious free-
dom that is specifically enshrined in the Bill of Rights as the First Amendment
to the Constitution – is sometimes used as a defence against accusations of mis-
leading advertising, which makes the regulation of advertising more difficult.
Within whatever constraints may apply, advertisers have devised many
modes of persuasion, often using numbers to reinforce their message. At the
naive end, we have claims in the style of “8 out of 10 cats who expressed a
preference said Wondercreme increased the length of their whiskers”. The
manufacturer will almost certainly be able to produce the results of a survey
that appears to show this – whether it would pass any test of peer review or
reproducibility is another matter. This kind of advertisement shows how you
can make a perfectly truthful statement about a ‘survey’ while falsely imply-
ing that something is scientifically proven. "Probably the best lager in the
world” is in this class, without even bothering with numbers.
The adverts in Figure 7.2.1 are imaginary, but they parody widely seen
advertising tricks. Simply quoting an irrelevant fact can make your bottled
FIGURE 7.2.1 Parodies of advertisement tropes.

water sound healthier than the equally fat-free stuff that comes out of the
tap. Few ‘miracle’ diet or exercise plans – even ones with some validity –
will work without additional major lifestyle changes (‘as part of your calory
controlled diet’ is the usual disclaimer). The third example, which arguably
crosses the line from ‘clever advertising’ to ‘outright scam’ is based on a story
from long ago (that might even be true) about a paid advert in a newspaper –
so the scammer took the risk that not enough people would bite to recoup the
costs. Today, mass distribution via social media or junk email costs virtually
nothing, so even a tiny ‘success’ rate can be profitable.
ACTIVITY
Questionable ads.
Ask students to bring in examples of advertisements they think are mis-
leading or might be misleading to some people, particularly if they rely
on numbers. Small groups then discuss and classify the methods used to
mislead.
Once a marketer has attracted a customer’s interest with a well-tailored

advertisement – they’re in the shop (or on the website) – there are further
tactics to seal the deal, stave off competition, and maximise profit.
The price on the sticker is the key number the customer sees, and their first
impression of how expensive the product is. Too high and it will scare off
customers. Too low and the seller is sacrificing profit. Techniques for mini-
mising the advertised price and obscuring the total cost of ownership include
the following:
• ‘Subscription services’ paying monthly versus lump sum one off pur-
chases have long been common for expensive things like cars (“Only $600
a month for 3 years”). This approach is now common for phone apps
too (“Only $4.99” – ‘per month’ is implied). It is spreading to physical
goods – for heated seats in your new car, for example – as a way to lower
the purchase price, while also providing a steady income stream for the
manufacturer.
• ‘The King Camp Gillette’ business model: give away the razors and then
charge a fortune for the blades. More modern examples include cheap
printers with expensive ink, or cheap coffee makers that use proprietary
coffee pods.
• ‘Optional’ extras: the advertised price gets you the bare-bones model, but
maybe you’ll regret not getting the more comfortable saddle for your bike,
the bigger battery for your electric car, more memory, and storage for your
phone/computer. The trick for the seller is to sell you these after you’ve
decided to buy the product at the sticker price.
In this kind of situation, taking time out to do simple calculations will usu-
ally let you compare the up front, annual, and total lifetime costs, leaving you
to consider which is more important to you at the time. If, for example, you
think that $200 of extra storage will make your new computer last 4 years
rather than 3 years, how will that change the annual cost?
Another way of making money beyond the listed price is to sell the cus-
tomer an extended warranty to insure against needing expensive repairs in
the future. As we asked in the Introduction, is it likely to save money? Here
it is important to consider when any breakdowns are most likely to occur.
Usually, this will follow the pattern of the ‘bathtub curve’ (see [7E] for back-
ground) shown in Figure 7.2.2.
The initial period when manufacturing faults show up is covered by the
standard guarantee. Wear and tear faults from repeated use increase after
many years, normally beyond the period of the extended warranty where the
likelihood of breakdown is at its minimum. The generosity of the standard/
statutory warranty varies between countries, and some extended warranties
will include cover for accidental damage that wouldn’t be covered otherwise
but, in general, it is more expensive to insure against risks that you can com-
fortably afford to cover than to pay when problems arise – insurance com-
panies make profits. This is another area in which bit of personal accounting
and modelling can help: how much do you typically spend, over a few years,
on repairing/replacing broken devices? How much would it cost to pay for
extended warranties on everything you buy?
FIGURE 7.2.2 Likelihood of breakdown of a typical consumer product.

7.3 Selling in politics

While most countries have laws against false or misleading advertising of
products and services, these rarely extend to claims and promises made by
politicians, who are free to mislead in whatever way they think they can get
away with. It is against the law for a manufacturer to falsely claim their car
will do 100 km on 1 litre of petrol, but not for a politician to say that Britain’s
economic performance over the last 10 years has been the best among the
OECD countries.
This situation is primarily justified as freedom of speech. It runs the same
risk of refutation by trusted fact-checkers with the associated loss of c redibility
and trust.
Claims for the future are also excused, sometimes even excusable, because
of the general unreliability of economic forecasting – a field described by one
distinguished economist as “designed to give astrology a good name”. None-
theless, evaluating political statements presents an ongoing challenge for the
mathematically literate in a very important domain.
In the UK, the major broadcasters (BBC, ITV, and Channel 4) have a remit
to present news and current affairs in an unbiased way that includes oppos-
ing views – one which they proudly claim to uphold. Print media, includ-
ing newspapers, have no such constraints. Nonetheless, the ‘leading papers’
work hard to achieve a reputation for presenting objective truth by sepa-
rating news reporting from opinion, following the (Manchester) Guardian’s
long-time editor, C.P. Scott (1921):
Comment is free but facts are sacred.
News media – originally newspapers but increasingly moving online – in the

United States that embrace this principle include The New York Times and
The Washington Post, The Age, and the Sydney Morning Herald in Australia.
We shall look at three UK publications. While the political viewpoint of each
is widely recognised, and clear in their comment sections, these organisa-
tions claim to present their news reporting as objective and unbiased. While
broadly true, there is some ‘leakage’ of opinion into news. Figure 7.3.1 shows
The Guardian The Times The Telegraph

15 April 2023 15 April 2023 15 April 2023
Further England nurses’ Nurses call 48-hour strike Nurses to target A&E in
strikes present ‘severe after rejecting RCN pay most extreme strike yet
challenge’ to NHS deal
FIGURE 7.3.1 UK news headlines – same news, different slant.

corresponding headlines from the three sources (one regarded as ‘progres-

sive’, two as ‘conservative’) in a dispute reflecting offers of pay increases to
nurses that are well below price inflation.
Note that The Guardian presents the strikes as a challenge for the health-
care system for which the (Conservative) government is responsible; The
Telegraph calls the strike ‘extreme’ and ‘targeted’ on emergency care, while
The Times suggests the offer comes from the trade union – which did indeed
recommend its members to accept the government’s offer as the best that
could be achieved. Other news sources make no attempt to separate news
from opinion.
This ‘leakage’ of opinion into news reporting, often far more blatant than
Figure 7.3.1, has led some to claim that there is no such thing as objec-
tive truth. The historical record, with its inevitable selection of facts, offers
evidence to support this view – the inclusion, or not, in history curricula
of slavery and its role in enabling Western economic success being a prime
example. Nonetheless, for literacy, mathematical, and, in general, to survive
and thrive, a search for objective truth through analysis of data and critical
thinking is central.
Ways of ‘improving’ an argument

While there is no reliable way of predicting the future in politics or econom-
ics, the rhetoric used by political parties for selling policies often relies on the
following:
• An account of past performance, which is verifiable though inevitably

complex.
• Comparisons between periods and/or between countries.
What are the essential components of these arguments and what ways of
misleading the listener or reader are common?
The choice of variables is, as always, crucial. For example, the favourite
indicator of the overall performance of an economy is GDP, the ‘gross domes-
tic product’ or, closely similar, the national income. However, since this says
nothing about where this income goes, it is a poor indicator from a personal
point of view. For most people the real median household income, adjusted
for inflation, is a more significant indicator of economic progress. (Avoid the
mean household income, which goes up even if all the benefit of an increasing
GDP goes to the very rich.)
Details in the selection of data can crucially strengthen or weaken a case –
for example, by choosing different start and end dates for the data used to
calculate summary statistics. This is illustrated by Figure 7.3.2, which shows
UK GDP from 1992 to 2021.
FIGURE 7.3.2 UK GDP 1992–2021 showing key events.

Source: Using open data from The World Bank: World Development Indicators 1/3/2023. [7F].
Often used for international comparisons.
ACTIVITY
GDP growth
Calculate the annual GDP growth rate that, if compounded over 10 years,
would produce GDP growth by a factor of about 2. Remember that growth is
measured in multiplicative terms (rates, ratios, percents), rather like interest
(see Chapter 9). This can be done using logarithms.
Alternatively, a spreadsheet can be used to quickly converge on the answer
using a guess-check–improve approach.
The appropriate formula for annual percentage rate R is as follows:
  loge G   end GDP

R = 100  e P  − 1 , where G = and P = period in years
  start GDP
For the growth factor of 2 over 10 years, this gives an annual compound
growth rate of 7.2% (to check calculate 1.07210). These figures match growth
in UK GDP from 1997, when a Labour government took over until the global
financial crisis of 2008. Those of a more conservative view would correctly
point out that by 2010, when the government changed, the growth since 1997
averaged only 3.7% (a factor of 1.6 over 13 years). In rebuttal, it could be

pointed out that for 2010–2021, the growth rate was only 2.1% (1.25 over
11 years) with no growth at all since 2015. These are very different economic
and political stories. Note that the graph provides a far more informative and
detailed picture of the situation than the summary statistics we just noted.
Playing with units and rates is another common device for giving a more
favourable impression – announcing the amount of new funding “over the
next three years” or “over the lifetime of this administration” (4–5 years,
depending on country) rather than on a standard annual basis makes the
government look more generous, until you think about it.
Reissuing announcements of funding that had already been promised pre-
viously is another common device. Sometimes this involves redirecting it to
another area. ‘Robbing Peter to pay Paul’ in this way is common.
Antidotes to distortions
The first line of defence against misleading political arguments is the media –
traditionally, the press and TV broadcasters. This has been disrupted by the
rise of the internet, where the distinction between ‘news’ and ‘social media’
commentary becomes blurred. Countries which genuinely aspire to having a
‘free press’ generally have a range of news providers with political analysts
who dissect the statements of government and opposition politicians. As we
have noted, most of them have a well-recognised political viewpoint, more
or less extreme, that needs to be taken into account. There is an even wider
range of choices on the internet, but only a minority aim for unbiased analy-
sis of political statements. The rest amounts to a flood of opinion, often very
strongly held and backed up by dis-information. So, again, in order to get a
reasonably objective overview, it is important to read the analysis in a range
of sources, including some that you regard as biased.
Some broadcasting organisations (including the BBC, ITV, and Channel 4
in the United Kingdom and NPR in the United States) have a public service
remit that requires them to hold politicians of all persuasions to account, ask-
ing the awkward questions. The answers – or frequently, the evasions to avoid
answering directly – give the listener some idea of where the weaknesses lie.
(That all political parties criticise the BBC as biased against them is some indi-
cation that it handles the public service remit pretty well.) This can, however,
lead to ‘bias by balance’ wherein a well-intentioned remit to include a range
of voices can result in minority views appearing more widely held than they
really are. For a long time ‘climate change deniers’ succeeded in this way – for
example, the BBC would give equal time to a leading climate scientist and a
retired politician who claimed the overwhelming evidence was inconclusive.
In the end a rigorous analysis of a political argument means going back
to trusted primary sources. Most countries have national statistical agencies
that not only produce reliable data on important areas but are also responsi-
ble for vetting government data to ensure that, while it may be selective, it is
not actively misleading.
While the majority of ‘think tanks’ have a political agenda – well recognised
if not admitted – there are some that have built a reputation for objectivity
without bias. The UK Institute for Fiscal Studies is an example – an enviable
reputation to achieve.
Confirmation bias and common misconceptions

It is important to remember that we all are, at least partially prone to ‘confir-
mation bias’, the ‘echo chamber’ effect – we seek sources that we are likely to
agree with, which provide comforting confirmation, particularly in areas where
we hold strong opinions. (The choice of examples in this chapter doubtless
reflects this.) To minimise this effect, it is important for informative discussions
to have people with opposing views. Personal vigilance is the only antidote.
In many cases, people’s strongly held misconceptions have been influenced
by popular culture – non-factual ‘memes’ that permeate news and entertain-
ment media, politics, and ‘what the bloke down the pub said’. Exploring and
analysing real-world data is one way of challenging this. The Gapminder
website [7G] offers several tools to support this. The headline feature is
the Worldview upgrader – a quiz with questions such as the one shown in
Figure 7.3.3 that play on widely held misconceptions. It won’t take long for
a smart person to spot the pattern of ‘common knowledge’ being wrong,
but this is justified by the survey data presented after each answer showing
FIGURE 7.3.3 Example question from Worldview upgrader and follow-up.

Source: Screenshot from Gapminder [7G] – modified to add key for black and white version.
how widely held the counterfactual views are. If using the quiz, it is impor-
tant to pay attention to this follow-up information, especially the part that
explains why the misconceptions can cause problems, even when the truth
seems more optimistic. Gapminder also offers a range of animated, interac-
tive graphs of demographic data, and supporting videos, which can be used
to explore issues raised by the quiz – these are highly interactive and colour-
ful, and best experienced by visiting the website. Sources of online data and
visualisation tools are discussed further in Chapter 10.
Some social media platforms exploit confirmation bias by targeting users
with material based on their previous browsing habits. In many ways, social
media have democratised the media, allowing people to share their views
without ‘the establishment’ – politicians, newspapers, and broadcasters –
acting as gatekeepers, curators, and censors. However, the resulting flood of
information and misinformation is too much for most people to sift through,
so the ‘new gatekeepers’ are the algorithms which recommend new material
to users based on their past browsing habits and other harvested personal
information. These algorithms are a crucial part of the big social media plat-
forms and – at best, assuming no more sinister motive – are designed with
the sole intention of keeping users on that platform by feeding them more of
what they like – there is no incentive to risk alienating someone by challeng-
ing their views. In any case, these algorithms work automatically without
thought or judgement. While this is also true of old-fashioned printed news-
papers, at least their biases are well known and their sometimes-conflicting
headlines visible to anybody browsing the newsagent’s shelves – they have
at least some incentive to be seen as honest, reputable journalists. The social
media giants are disinclined to take decisive action against misinformation
not because they have some evil plan, but because their entire business model
relies on automatically processing vast quantities of material contributed by
customers, and any meaningful level of human curation would be prohibi-
tively expensive. They say they are not pretending to be journalists – they are
simply connecting you with your newly expanded circle of ‘friends’.
ACTIVITY
Discuss how social media have affected the informative discussion of issues.
On bullshit
We noted earlier that there is an approach to persuasion that doesn’t worry
about the consequences of telling lies. The method is to bury the listener or
reader in bullshit – a string of assertions, sounding more or less plausible,
that support whatever argument the speaker wants to make. Truth is irrele-
vant. In Frankfurt (2005), the Princeton philosopher gives a readable account
of bullshit and how to spot it. Long familiar in social groups with drinks in
hand, in recent years some charismatic politicians have found that it works
to advance their careers, at least for a while. So recognising the difference
between lies and bullshit has become an essential skill for assessing ‘informa-
tion’, in our politics as elsewhere. Being mathematically literate can help.
Acknowledgements
Figure 7.3.2 was created using open data from The World Bank: World
Development Indicators 1/3/2023. [7F].
Figure 7.3.3 is a screenshot from the World View Updater website – free
material from www.gapminder.org [7G].
References
Frankfurt, H. G. (2005). On bullshit. Princeton University Press.

purposes:
[7A] The Secret World of Fake Online Reviews – The Guardian (22 April
2023)
https://www.theguardian.com/money/2023/apr/22/it-can-be-incredibly-
profitable-the-secret-world-of-fake-online-reviews
[7B] Best Washing Machines to Buy in Australia – Choice
https://www.choice.com.au/home-and-living/laundry-and-cleaning/
washing-machines/review-and-compare/washing-machines
[7C] Decathlon: The Art of Scoring Points – NRICH
https://nrich.maths.org/8346
[7D] Spreadsheet to Compare Total Cost of Phone Ownership – mathlit.org
https://ltml.mathlit.org/7D
[7E] Bathtub Curve – Wikipedia
https://en.wikipedia.org/wiki/Bathtub_curve
[7F] World Development Indicators – The World Bank
https://databank.worldbank.org/source/world-development-
indicators
[7G] Worldview Upgrader – Gapminder
8
EQUALITY AND INEQUALITY
Inequality of wealth and income is a topic that excites the interest of

people, particularly young people whose relative position has declined
in recent decades. As the word implies, it is about comparisons between
different groups at different times, offering an opportunity for discus-
sions of important social and political issues where numbers are central.
But, as with any complex field, there are many issues to consider. When,
for example, should we include wealth, where the disparities are even
larger, along with income? Finding appropriate data and choosing it to
tell a story is at the heart of mathematical literacy. Here more than in
most areas, the choice a person makes in presenting a case will depend
on the message they want to get across – which is interesting in itself.
So this is very much a ‘know about’ chapter, where critical thinking is
central.
For example, a popular measure of inequality is the ‘wage ratio’ be-
tween the chief executives of large companies and the median wage of
their workers. In the United States, this rose from about 20 in 1950
through 120 in 2000 to 650 in 2022. An analysis noted that “on the
first workday of 2023, CEOs will make more than the average annual
pay for all US workers”. (Quick check: If the average working year is
260 days per year, the wage ratio needs to be 260 for the first day pay to
equal average annual pay.) The ratio has increased over the last 50 years
in other countries, too, from single digits to around 100 in the UK, in
Sweden to 50. There is no evidence that this reflects improved company
performance.
DOI: 10.4324/9781003303503-9
160 Equality and inequality
ACTIVITY
Is inequality ‘a bad thing’? If so, in what ways. Small groups discuss the advan-
tages and disadvantage of inequality of income, as represented by the ratio of
CEO to median worker income. What ratio seems appropriate?
This data represents a surprisingly extreme contrast in incomes, yet com-

parisons of wealth yield larger ratios. Is there a way to get a picture that is
both broader and more balanced so as to inform discussions – in the class-
room or elsewhere?
8.1 A historical analysis of wealth and income

The research on these issues by the distinguished economic historian Thomas
Piketty is summarised in his book A Brief History of Equality (Piketty,
2022) – a very readable account for the non-expert mathematically literate.
Here we outline some of the issues by presenting just a few of the compara-
tive graphs from the book – all are available from Thomas Piketty’s website
[8A]. This website offers a rich source of information for small group and
class discussions of the many important issues involved.
Figure 8.1.1 shows how both the total wealth of the world and the popu-
lation have grown over the last 300 years. (They were relatively stagnant
through the previous centuries.)
FIGURE 8.1.1 World population and average income 1700–2020.

Source: Adapted from Piketty (2022) [8A].
Equality and inequality 161
Briefly, world population and average national income per person both in-
creased more than tenfold between 1700 and 2020 – from about 600 million
inhabitants in 1700 to over 7 billion in 2020, while income per person (ex-
pressed in 2020 euros with ‘purchasing power parity’) increased from barely
€80 per month per person in 1700 to €1000 per month per person in 2020.
This corresponds in both cases to an average annual growth rate of about
0.8%, compounded over 320 years. Note this means that total world income
(GDP) increased at an average rate of about 1.6%. (Mathematical literacy
question: 0.8% more people each year each with 0.8% more income – why
is adding the rates reasonable here?)
So we have got richer. But is it really that simple? Does ‘2020 purchas-
ing power parity’ mean anything before the industrial revolution, or at a
time when you could go to prison for being in debt. Piketty (2022) outlines
how those issues have been investigated in a comprehensive programme of
historical analysis of prices, wages, wealth, and other variables. Equally,
he recognises that in comparing such different ‘worlds’, the results are only
qualitatively meaningful. However, his conclusion that the average person is
economically much better off than 300 years ago is surely plausible.
Average income can be a useful measure for comparisons between coun-
tries and different periods, but it obscures inequality between people at
different levels of income within each country. Piketty often tracks the in-
come and the wealth of three groups: the top 10%, the middle 40%, and
the bottom 50%, comparing them over time and relating them to political
and economic changes in the periods concerned. He observes that, while all
groups have become more prosperous, the general trend is for increases in
prosperity to go mainly to the already wealthy – a trend that has only been
contained by active government intervention, mainly through redistributive
tax systems.
This issue is vividly illustrated by comparing the post–World War II ‘egali-
tarian’ period 1945–1980 with the more individualistic ‘neoliberal’ econom-
ics that was introduced in the 1980s by Fraser (Australia), Thatcher (UK),
and Reagan (USA). This approach is built on the economic ideas of Hayek,
Friedman, and others in the ‘University of Chicago School’. They argued
that individual entrepreneurship aided by low tax rates would produce faster
growth (GDP) whose benefits would “trickle down” to raise the incomes of
everybody. “All boats rise together with the tide of increasing prosperity”. It
has not turned out that way.
Figures 8.1.2 and 8.1.3 show how the sharing of wealth and income
between the classes changed between 1900 and 2020. In Europe as in the
United States, we see that between 1914 and 1980 a steep decline in the
share of the richest 10% in total private property (real estate, business and
financial assets, net of debt) to the benefit principally of the middle 40%.
This movement is then partially reversed between 1980 and 2020, notably
FIGURE 8.1.2 Wealth in Europe and the United States, 1900–2020: The birth and
fragility of a patrimonial middle class.
in the United States. (“Europe” here is an average of the data from France,
Germany, Sweden, and Britain.) Note that these percent shares are not per
person. In Europe in 2020, for example, the top 10% together had well more
than 8 times the wealth of all the people in the bottom 50% together, so in-
dividuals at the top had more than 40 times the wealth of individuals at the
bottom. In the United States, the situation, as the graph shows, the ratio is
more extreme.
FIGURE 8.1.3 Income inequality: Europe and the United States, 1900–2020.
Figure 8.1.3 shows the share in national income of the richest 10% ver-
sus the poorest 50% and how the pattern of income changed over the same
period. In Europe, income inequality has started to rise again since 1980,
although remaining at levels clearly lower than those of 1900–1910. The in-
crease in inequality has been much greater in the United States. In both cases,
inequality has remained high: the richest 10%, though five times fewer, still
receive a share of total income much larger than the poorest 50% receive –
on average about 15 times greater per person in the United States, 8 times in
Europe, with huge disparities between individuals in each case, particularly
within the top 10%.
How far does this reflect the pattern of taxation?

Here we focus on the United States where, from 1915 to 1980, the tax sys-
tem was highly progressive, in the sense that the effective tax rate paid by the
highest income groups (taxes of every type included, then expressed as a per-
centage of pre-tax income) was substantially larger than the average effective
tax rate paid by the total population (and particularly by the bottom 50%
incomes). Since 1980, the tax system has been much less progressive, with
little differences in effective tax rates across groups, as shown in Figure 8.1.4.
In 2020, people in all groups paid about 30% of their income in taxes.
However, the promised resurgence of growth following the cut in top tax
rates, predicted by the neoliberal economists and their often-wealthy sup-
porters, did not occur. In the United States, the top marginal tax rate applied
to the highest incomes dropped from 72% to 35% and the growth rate of
per capita national income dropped from 2.2% to 1.1% per year over the
FIGURE 8.1.4 Effective tax rates and progressivity in the United States 1910–2020.
TABLE 8.1.5 Growth and progressive taxation in the United States, 1870–2020
Period 1870–1910 1910–1950 1950–1990 1990–2020
Top tax rate 0.8% 55% 72% 35%

Growth rate 1.8% 2.1% 2.2% 1.1%
Source: Data from Piketty (2022) [8A].
same period – as shown in Table 8.1.5. And, as we saw above, far from the
predicted trickling down of riches so that “all boats rise together with the
tide”, the inequality of distribution of wealth and income surged upwards.
There were similar but less extreme effects in Europe, where the diverse tax
systems make comparisons more complicated.
ACTIVITY
A critical discussion of the pattern of wealth and income data from this section.
Students can be encouraged to look for other data, from piketty.pse.ens.fr/
equality and other sources, to be used in their preparation.
8.2 Social effects of inequality

What are the effects of inequality? Is inequality ‘a bad thing’? After all it is
broadly favoured by the wealthier, who don’t like to pay a lot of tax (who
does?), and generally accepted as a fact of life by those who have little but
‘somehow get by’.
In their book The Spirit Level: Why More Equal Societies Almost Always
Do Better, Richard Wilkinson and Kate Pickett (2009) present data, sum-
marised in Figure 8.2.1 for rich (OECD) countries, showing that health and
social problems are worse (i.e. higher level of problems on the graph) in socie-
ties with higher income inequality. The index of health and social problems
combines measures of life expectancy, infant mortality, teenage pregnancies,
obesity, mental illness, homicides, imprisonment rates, social mobility, and
trust between citizens. The authors present detailed data showing correlations
with inequality – some stronger than others, but all in the same direction for
each of these factors. This and subsequent research data can be explored on
the Equality Trust website [8B], a rich source of information on this issue.
Their measure of income inequality is the ratio of the averages of high-
est 20% of incomes to the lowest 20%; this varies from below 4 in Japan to
nearly 10 in the United States. They also show that the index does not corre-
late with the average income of the country – Greece and Ireland, for example,
are close in income inequality despite a 50% difference in average income.
FIGURE 8.2.1 Index of health and social problems versus income inequality.
Source: Adapted from Wilkinson and Pickett (2009) [8B].
ACTIVITY
Use the data from The Spirit Level [8B] to inform discussions of the social effects
of income inequality on different income groups. Ask each student to choose
one graph and explain its implications for discussion in the group.
ACTIVITY
Examine the differences between the measures of income inequality that are
used by Wilkinson and Pickett (e.g. Figure 8.2.1) and Piketty (Figure 8.1.3) and
discuss strengths and weaknesses.
Methodological issues
In considering this data and the dramatic inferences that it suggests, there
are some important issues of principle that a mathematically literate person
may well be considering. The data shows correlations between measures of
inequality and social outcomes – but correlation does not necessarily imply
causation, nor say in which direction that might flow. Since it is normally
impossible to do controlled experiments on societal variables, one relies on
observing the outcomes of changes that arise naturally or through the effects
of policy decisions.
For example, tobacco companies argued for a long time that the clear cor-
relation between smoking and lung cancer might be due to some common
causative factor – a tendency to be anxious, perhaps. This argument lost
credibility when changes in smoking habits were matched by reduced cancer
deaths – as well as by experiments that revealed the causative mechanisms.
Wilkinson and Pickett devote later chapters of The Spirit Level to such
methodological issues, responding in advance to legitimate questions of
this kind. A mathematically literate person may have questions about the
mathematical modelling of social constructs (e.g. how the measures are
created – for more on this see Chapter 12), the quality and comparability
of data over time and between countries, the logical premises of the argu-
ments (e.g. correlation/causation as above), and the critical thinking about
the conclusions.
8.3 Approaches to reducing inequality

The Spirit Level also looks at the springs of inequality in different societies
and the kinds of change in policy and practice that might reduce it. They
point, for example, to the potential of ‘pre-distribution’ through reducing
inequality of income (as in Japan) rather than relying on ‘redistribution’
through taxation. They advocate a range of ‘levers’ to reduce inequality, in-
cluding more progressive taxation as in the past, a change in corporate law
from its focus on shareholder value to include other stakeholders (e.g. lead-
ing to better paid jobs for employees), and a social ‘safety net’ that supports
a decent standard of living.
Complementing this broad look at various mechanisms, Piketty describes
a specific approach. While his historical analysis has been widely accepted,
his proposed way forward is much more controversial. Nevertheless, it is a
useful starting point for discussions of the important issues raised.
Piketty’s model for participatory socialism treats wealth and income as
comparably important. It is based on redistribution through two channels:
the circulation of property between generations – a minimum inheritance to
be allocated to each young adult at 25 years of age – and a basic income for
all. The effect of the former is illustrated in Figure 8.3.1.
In 2020, the share of the poorest 50% in total inheritance was 6% in
Europe (Britain–France–Sweden average), with 39% for the middle 40%
and 55% for the richest 10%. After implementation of a minimum inherit-
ance equal to 60% of average wealth, financed by a progressive tax on both
FIGURE 8.3.1 The distribution of inherited wealth, current, and projected.

wealth and inheritance, the shares would be 36%, 45%, and 19% for the
three groups. Given the 50:40:10 ratio of population in the three groups,
these shares correspond to ratios of about 7:11:19 for the average individual
in each group. As now, there would be wider disparity within the top 10%
than in other groups.
The proposed tax system to finance this redistribution includes a progres-
sive tax on property (an annual tax plus inheritance tax) funding the capital
endowment for young adults, together with a progressive (roughly loga-
rithmic) tax on income to fund both the basic income and public services –
health, education, pensions, unemployment, energy, and so on.
In the example shown in Table 8.3.2, the progressive property tax amounts
to about 5% of national income – enough to fund a capital endowment of
about 60% of average net wealth to be allocated to each young adult at
25 years of age. The progressive income tax raises about 45% of national
income – enough to fund an annual basic income of about 60% of average
after-tax income, costing about 5% of national income, and the public ser-
vices of a social and ecological state costing about 40% of national income.
ACTIVITY
In the light of the data presented in this section, discuss the pros and cons of
the Piketty model from both ‘fairness’ and ‘economic’ points of view, as well
as ‘levers’ identified by Wilkinson and Pickett.
TABLE 8.3.2 Piketty’s model for circulation of property and progressive taxation
Progressive tax on property Progressive tax on income

Funding capital endowment for each young adult Funding basic income and
services
Multiple of Annual tax Inheritance Multiple of Effective

average wealth tax average income tax rate
0.5 0.1% 5% 0.5 5%
2 1% 10% 2 10%
5 2% 50% 5 50%
10 5% 60% 10 60%
100 10% 70% 100 70%
1000 60% 80% 1000 80%
10 000 90% 90% 10 000 90%
Source: Data from Piketty (2022) [8A].
While this model may seem extreme, and the inheritance aspects are un-
precedented, the top income tax rates are comparable to those operating
in the United Kingdom and the United States in much of the 1950s when
growth was faster and inequality was lower than it has developed since 1990
(Figures 8.1.3 and 8.1.4 show the US data). The top rate of income tax in the
UK was around 90% in the 1950s but has dropped to about 45%. Inequality
in the UK increased sharply during the 1980s and is now among the great-
est among the OECD countries (see Figure 8.2.1). Understanding issues like
these is an important goal of mathematical literacy for informed citizenship.
In a democracy, there will always be different opinions on the best way for-
ward, but making good decisions depends on knowing the broad parameters
under which your society is operating.
References
Piketty, T. (2022). A brief history of equality. Harvard.
Wilkinson, R., & Pickett, K. (2009). The spirit level. Penguin.
Acknowledgements
Figure 8.2.1 is based on a figure taken from Wilkinson and Pickett (2009);
the original is available in a slideshow available via link [8B] below. All of
the other figures are adapted from Piketty (2022) and appear courtesy of the
World Inequality Lab. The originals can be found at link [8A] below.

purposes:
[8A] A Brief History of Equality – Thomas Piketty

http://piketty.pse.ens.fr/fr/equality
[8B] The Spirit Level – Equality Trust
https://equalitytrust.org.uk/resources/the-spirit-level
9
YOUR MONEY IN YOUR LIFE
When thinking about using mathematics in everyday life, money is the first domain
that most people think of. Control of your finances has always been an important
ingredient in human happiness. It is also worth being more sophisticated than
Mr. Micawber’s advice to David Copperfield, in the context of Victorian England:
Annual income 20 pounds, annual expenditure 19 pounds, nineteen and six,

result happiness. Annual income 20 pounds, annual expenditure 20 pounds
ought and six, result misery.
Charles Dickens (1850). David Copperfield, Chapter 12
The need to “reconcile my net income with my gross habits” (attributed to

‘30s film star Errol Flynn) encapsulates an important strategy for everyone.
But it is not specific enough to be much help on its own. What are the issues?
What questions should we ask ourselves in the modern world?
This chapter aims to bring out underlying principles that enable people to get a
clear view of where they are financially, and the options that are available to them
on different timescales – but without their spending more time on details than
their circumstances and inclinations demand. It is written on the assumption that
those involved will live and work in a developed country and will be empowered
by a deeper understanding of issues such as debit, credit, investments, consumer
choices, and other aspects of personal financial management. While the principles
that are discussed are general, some of the specific situations we discuss will be
relevant only to some people, and only in the future. As ‘believable’ rather than
‘action’ problems, they are an important part of mathematical literacy. Indeed,
the issues discussed are mainly the concern of adults, so the examples used may
be rather different to those that are relevant for young teenagers.
DOI: 10.4324/9781003303503-10
Your money in your life 171
We do not aim to provide a course on financial literacy, which in recent

years has become an important topic in schools with many curriculum ma-
terials developed (see, for example, Sawatzki, 2016). Financial literacy has
been assessed in every PISA cycle from 2012, so basic data on the financial
knowledge of 15 year olds around the world is accumulating. The framework
for the 2022 PISA financial assessment [9G] provides an overview of find-
ings, and describes how the financial services directed to young people have
changed. More services are now available, even to children as young as 5 in
some countries, and new products are often aimed at young people, including
‘buy now, pay later’ purchasing with high penalties for missed payments.
Financial literacy courses are also available online and there is an ever-
expanding range of apps that enable people to keep track both of their financial
transactions and of the financial environment in which they operate. However,
using the rich resources they offer tends to take a lot of time – itself a precious
commodity. One definition of being rich is ‘shorter of time than money’.
In getting to grips with the essentials of finance, as always with problem-
solving, identifying the important variables is the starting point, moving on
to explore the relationships between them. Here the key variables include
income, expenditure of various kinds, capital, debt, growth of various kinds,
and inflation; others will emerge in the course of the chapter. Overarching
variables, and associated questions, include the following:
• Timescale: How are things financially at the moment? How urgent is it for
you to get more money? How might you do it? Would you be willing to wait
and study for a while, or work for lower pay for the benefit of experience, so
as to get more in the end? If so, is waiting a few years OK?
• Risk: In financial matters do you accept, even enjoy the risk element, or,
conversely, are you ‘risk-averse’? Financial decisions, particularly long-term
strategic ones, have some uncertainty in outcome – hence the standard
phrase accompanying financial advice “may go down as well as up. Past
results are no guarantee of future performance”.
Questions like these are the basis of what follows – and of the learning
activities we suggest. Money can be a sensitive subject, with people at very
different levels of wealth, income, or debt in the same class. The activities
here are structured to avoid direct comparisons that may be uncomfortable
for some students. How they are used is, as always, a matter for the teacher.
Most issues in this chapter will be familiar to most teacher educators,
many of them to their students, fewer to the school students they will teach
who will have other experiences and perspectives. So the challenge is to find
the most effective teaching approach to develop robust learning at each level.
Discussion of the processes of financial literacy is central; on the content
the challenge is to find an appropriate balance between a quick review (as
here) and careful direct explanation of all the details. As usual, the suggested
172 Your money in your life
activities are structured to promote discussions that elicit the key issues, guid-
ing the decisions that can help improve quality of life in the broad sense.
9.1 Knowing where you are

This is about strategy on the shortest timescale – and the least ambiguous
issue, though not entirely without complexities. It offers a simple example of
working at different levels of detail. Channelling Mr. Micawber, these ques-
tions will usually stimulate a lively discussion with a range of responses:
• Are we spending more than our income?

• Are we spending it on the things we really want? How can we check on this?
A detailed approach
A commonly recommended approach is detailed budgeting – keep a note of every-
thing you spend in a week (or a month) and compare it to what you get in. There
are many ways of doing this: from a simple notebook to a spreadsheet or an app
designed for the purpose. The process is somewhat laborious, but the detailed pic-
ture it provides is crucial to gaining mastery over one’s finances. To be really use-
ful, this approach needs expenditure to be broken down under various categories.
Figure 9.1.1 shows a simple spreadsheet layout; the formulae involved can
be seen by downloading the spreadsheet [9A].
FIGURE 9.1.1 A spreadsheet for detailed accounting.

Source: Spreadsheet available online [9A].
The detailed information, particularly the totals for each category,

give a picture of where the money is going; this may be compared with an
expenditure plan – the ‘budget’ – or may inform the process of m aking
one.
ACTIVITY
Everyone in the class tries this for a week, or a month, choosing their own
categories. In the subsequent discussion, the teacher might ask questions like
these. How did the process work out for you? Did you carry it through? Did
you learn anything useful about your finances? I don’t want numbers; just
your experiences of the process.
To repeat: it is clearly important to avoid surfacing actual numbers for
income or expenditure – some members of the class will have much more
money than others. The sample spreadsheet offers a number-free pie chart
(Figure 9.1.2) showing the spending broken down into categories, which
could be a less sensitive subject for discussions.
FIGURE 9.1.2 Number-free pie chart for comparison of expenditure.

Source: Spreadsheet available online [9A].
The big picture

This may be the point to present a question of strategy for discussion. Is there
a way to cut through this complexity? Can you find a simple way of knowing
where you are, month by month, without tracking the details? Ideas will
surely emerge, including some of the following.
Historically, when all payments and earnings were in cash, you could add
up all your money once a month (or once a week) and compare it to the
previous month. This was a straightforward route to the big picture of your
current finances. Adding a bank account meant the change in the balance
over the month had to be included, as well as payments made in other ways
such as cheque or lay-buy or hire purchase or debts.
Now people generally have many different accounts, but with technol-
ogy you can check all your accounts online on the same day. The details
are changing rapidly as technology evolves, but the basics of income versus
expenditure and the cost of credit remain, and most online or mobile pay-
ment services ultimately link to a credit card or bank account. (The emerging,
and controversial, world of cryptocurrency might see that disrupted in the
future.)
One issue that may arise is how you treat items bought on a credit card.
Conservatively, as expenditure on the day that each item was incurred,
even though you won’t pay them for up to six weeks – or much longer if
you don’t pay off the balance each month. How do you account for future
expenditure incurred but not yet made, such as direct bank transfers or
other periodic payments? That takes the conversation into issues such as
debt and interest that go beyond ‘knowing where you are’. We come to
them later.
As things begin to look more complicated, keeping a complete record
of all your expenditure may sound like a simplification but, as we saw, it
involves a lot of detailed work – always something to avoid unless it is essen-
tial. There is a technology-based alternative that many people use – to make
all your purchases through one account, for example, a debit card or by
electronic funds transfer. The early assumption in the COVID-19 pandemic,
that transmission was mainly through touching infected surfaces, pushed this
change forward, especially with the advent of ‘contactless’ payments. When
combined with the bank statement, which gives income and direct transfer
payments, this can make seeing the big picture of month-to-month changes
easier.
ACTIVITY
Compare detailed tracking with a big picture approach. Keeping a detailed

record of expenditure is time-consuming. How often, and when, is it worth
the effort?
In practice, many people who are living comfortably within their income
don’t track their expenditure in detail but keep an eye on ‘the big picture’
from time to time. However, when circumstances change – new place or new
job, for example – it makes sense to monitor things more closely.
ACTIVITY
Discuss the challenges that arose in this discussion – mainly identifying

variables and then finding, organising, and processing information. What have
been the challenges in thinking through your current finances? After all, the
mathematical concepts and techniques are simple – essentially just arithmetic.
This nicely illustrates the differences between strategic, tactical, and tech-
nical elements in the overall difficulty of a task – and how, in contrast to
most school mathematics which is technically demanding for the student,
the first two are often the main challenges in developing mathematical
literacy.
Capital versus cash flow

This is a good point to surface the important distinction between these basic
concepts. Capital is how much money you have at a point in time; the units
are money: £, $, €, and so on. Cash flow, income – expenditure, is a rate – the
clue is the word ‘flow’; the units are money/time: £ per year, $ per month,
for example.
Detailed record-keeping notes every ‘in’ or ‘out’ event over a period.
The big picture approach calculates cash flow by measuring capital at two
different times and dividing by the interval, avoiding having to calculate
everything directly in detail.
ACTIVITY
In preparation for a discussion, ask each student to calculate their cash flow
in a recent month by both methods, detailed versus big picture. Then they
should compute the percentage difference between the two numbers – if any.
Later, small groups discuss the advantages and disadvantages of the detailed
budgeting versus big picture approach. They report on findings of differences,
and evaluate accuracy against effort.
Credit versus debt

This important distinction is mainly one of attitude – in both cases someone
owes money, often you! Credit is a particularly slippery term. Credit is often
used for money we have received on the agreed understanding that we will
repay the ‘loan’ later in a well-specified way. Your credit rating reflects a
history of how well you met these promised repayments in the past. A good
credit rating will be used as a guide in future by people considering giving
you credit for something you want.
But on some occasions, you are the creditor. When your bank account is
‘in credit’, meaning a positive balance, you are lending money to the bank –
on the understanding that you can recover it on agreed terms – immediately
you want to or after a specific period.
Debt is a more general term for money you owe, and unless it is a loan from a
friend or family, it is likely to incur regular interest payments at a specified rate.
It may or may not have a specific date for ‘paying off the debt’. Failure to make
the payments will have consequences – financial, or sometimes more serious.
Ironically, credit cards offer free credit up to the monthly repayment date
when they essentially become ‘debt cards’ – with a high rate of interest on the
balance. In contrast, ‘debit cards’ draw directly from your bank account; they
will only pay out if you have money in your account, so no interest is involved.
9.2 Where do you want to go? The issue of values

The obvious answer is “I would like as much money as possible” (“within
the law” is sometimes added or implied), but this is rarely true – most people
have limits to what they are prepared to do for money. It leads on to other
questions. Financial literacy courses often present this issue as Need versus
Want, but this too is simplistic. Answers to such questions are value-based
and personal, often influenced by social pressures as well as circumstances.
What is a ‘need’?
Let’s start with the shortest timescale: current needs. Beyond the basics of
food, shelter, and safety, many things over the last century have started as
novelties but later come to be regarded as necessities: a car, a machine to wash
and dry clothes, a dishwasher, a vacuum cleaner, a microwave oven – all these
are practical aids to making life easier. While they are not, strictly speaking,
essential to life, the pace of modern society – especially the expectation that
mothers as well as fathers will have jobs – makes them hard to do without.
This concept of ‘necessity’ is in continuous flux: the ‘landline’ telephone,
radio, and television started as novelties for the rich, became near-necessities
but are now being replaced for some people by the smartphone with internet
access. Social changes have made gym subscriptions, subscriptions to sports
TV/video channels, and fast fashion feel essential for many. The rise of the
‘nail bar’ is a vivid example of something moving over a few decades from
invention to being a ‘want’ and then, for some, a ‘need’. There are also activi-
ties that might be regarded as ‘needs’, such as playing sport, learning music,
going to the cinema, travel, or going out with friends.
ACTIVITY
Introduce by asking: Beyond enough money to stay dry, warm, and fed, what
are the needs on which most people agree? Then,
• ask each person to list their needs and their wants, and then to rank order
items in each set;
• compare lists within a group, identify differences, and try to elicit underly-
ing factors – and any principles that seem to emerge; and
• ask a spokesperson for each group to present their lists and answer questions.
How far down each rank-ordered list a person can go is determined by how
much money they have (or can borrow – we will return to that). If ‘getting as
much money as possible’ is your overriding goal, it implies that you stick to
your ‘needs’ list – excluding everything else. The discussion will usually show
that this is rare.
Thinking longer term

In thinking on a longer timescale, the financial scale is correspondingly larger.
Currently, the order of magnitude of the lifetime earnings of a typical person
who works for 40 years is around $1 000 000 after taxes, but with big vari-
ations depending on background, career choices, support, and luck. To take
an obvious example, the decision on whether to go to university or to get a
job straight from school is a complex trade-off between
• immediate income versus the prospect of earning more after a deferred start;
• higher salaries available to many graduates versus the long-term costs of
paying off student loans;
• the benefits of the university experience; and
• a potentially more interesting job.
Estimates of increased lifetime earnings with a bachelor’s degree – which

some say is about 20% in real terms – are dependent on assumptions about
the future of jobs and the economy that are uncertain, to say the least.
ACTIVITY
In small groups, explore the estimates of income for different jobs, bearing in
mind how long or expensive the training is. Discuss trade-offs such as those listed
above. Some students may be interested to find data to improve their estimates.
Values and goals

The questions a good financial adviser asks a new client will be designed to
understand the client’s values and priorities in some depth. The first step is to
ask the client to discuss what they want to achieve, including things like secu-
rity, freedom, job satisfaction, time for family and friends, a sense of accom-
plishment, making a difference to other people’s lives, and work–life balance.
The adviser uses this information to develop a sense of the client’s priorities.
The next step is to turn these values into specific goals, with target dates,
the money needed to reach them, and – less obvious, perhaps – the satisfac-
tion that achieving each is likely to bring. Together, these form the basis of a
‘financial road map’ to guide planning (see, e.g., Bachrach, 2000). This is not
a ‘one off’ activity but is reviewed regularly as circumstances or values change.
ACTIVITY
Discuss the value of a process that claims to measure client values and financial
goals.
Setting financial goals in one’s youth may well seem unrealistic but, in reality,
this is just the right time – essentially because of the power of the exponential.
It seems obvious that an individual with clear goals is likely to modify their
behaviour over time, and very likely their financial habits, in pursuit of them.
9.3 Growth and decay

In the section on epidemics in Chapter 4, we saw the increasingly rapid growth
that characterises exponential behaviour – and how difficult it is for us to
grasp this behaviour intuitively. Financial thinking reflects this, in both direc-
tions. A modest amount of capital invested at an apparently low interest rate,
compounding over many years, grows into something substantial. The power
of compound interest to multiply capital over long periods is a concept at the
core of good financial planning. Getting to understand exponential behaviour
is core mathematics and central to many aspects of mathematical literacy.
Exponential growth and money

A spreadsheet like that in Figure 9.3.1 is a good way to see how the exponen-
tial grows, row by row, faster and faster. The graph shows how the growth is
mostly towards the end, particularly over many ‘doubling times’. The famous
‘grains of rice on a chess board’ problem, described in Chapter 2, is mirrored
FIGURE 9.3.1 Doubling time investigation.

Source: Spreadsheet available online [9B].
in the ‘100% per interval’ column in Figure 9.3.1. Other columns represent
growth with various interest rates – which are usually much larger for debts
than for savings! It may be worth spending time discussing the implications
of how the result changes with the interest rate. Explore the concept of
‘doubling time’ through examples, as in the figure. The live spreadsheet
used for the figure can be downloaded from link [9B].
ACTIVITY
Adam is 25. He is looking to save money for a pension starting with £100.
He has been offered an investment paying 3% per year interest. This can be
investigated with a simple spreadsheet (Figure 9.3.2 – link [9C])
• If he leaves the interest on his initial £100 to ‘compound’ in the investment,

how much will it be worth at age 65? (£326)
• If he continues to add £100 each year, what will the total be? (around
£8000)
• How does this compare with the total he has invested? (£4000, i.e.
doubled)
FIGURE 9.3.2 Simple savings calculator.

Source: Spreadsheet available online [9C].
More challenging mathematics

Some groups will be able to work out the doubling time in the general case
by solving, for n:
n
 i  ln2
 1 +  = 2 for n, giving n =
100   i 
ln  1 + 
 100 
This solution is included at the bottom of each column in Figure 9.3.1.

There is a useful ‘rule of thumb’ which says that, for small interest rates,
doubling time in years is approximately 70 divided by the annual percentage
rate of interest. This is derived from the solution above by taking the first
term from the Taylor’s series expansion of the denominator:
 i  i i i ln2 100 ln2 70

ln  1 + = + + ≈ , so n ≈ ≈ ≈
 100  100 1002 100  i  i i
 
100
(Note that this same formula was derived differently in Section 4.4 in finding
the time for cases to double in an epidemic. The exponential function links
both contexts.) To go into the mathematics more deeply, this situation provides
an opportunity to relate the discrete mathematics of stepwise growth to that
of continuous change (sometimes offered by US banks as “instant interest”).
How do difference equations relate to the associated differential equation?
ACTIVITY
Investigate what interest rate per month/week/day/instantly corresponds to

(say) 10% per year.
Inflation
Figure 9.3.3 shows UK prices since 1820 that are equivalent to £100 in 2020.
Note that over many years in the nineteenth century, the prices of goods re-
mained much the same but since World War I that has changed profoundly,
with prices rising year-by-year more or less rapidly. Note the greater clarity
of the logarithmic scale when things change by orders of magnitude, as here
where about £1 in 1820 would buy as much as £100 about 200 years later.
The issues involved in comparing prices over such a long period when buying
patterns have changed are discussed in O’Donoghue et al. (2004).
Change in the value of money over time has led to the distinction between
the cash value of something and its real value when allowance is made for
FIGURE 9.3.3 Equivalent prices in UK, 1820–2020.

Source: Data from O’Donohugh et al. (2004) and UK Office for National Statistics [9J].
inflation. So in the previous activity, if the inflation rate was 3% per year
over the 40 years, the investment at 3% interest would be unchanged in real
value. It has only held its value. In this artificial example, the growth from
interest earned has been matched by the decay in real value from inflation.
On shorter timescales, inflation is often measured by comparing the price
year-by-year of a ‘basket’ of goods that reflects what a typical consumer buys –
it is essentially the gradient of the curve in Figure 9.3.3a; the percentage increase
is the consumer price index, shown for the same data smoothed from year-to-
year in Figure 9.3.4.
FIGURE 9.3.4 Consumer price index of inflation, 1820–2020.

Source: Data from O’Donohugh et al. (2004) and UK Office for National Statistics [9J].
Economists argue about the ideal rate of inflation – perhaps surprisingly,

it is generally agreed that zero is not optimal. Why? Mainly to avoid com-
ing close to deflation where prices fall. This has negative consequences for
economic activity because it increases the real value of debt and discourages
spending, because consumers expect prices to keep falling.
The growth of debt

Exponential growth applies equally to debt – and generally at a much higher
rate than investment can return. The interest rates charged on credit card bal-
ances after the monthly payment date vary widely, but typical rates can be
as high as 25% per year or even more. At this rate the spreadsheet in Figure
9.3.1 shows that without any repayment, the money owed doubles in about
3 years. Even the approximation 70i = 70
25 = 2.8 created for ‘small’ interest rates
gives a salutary idea of the doubling time. Credit cards are convenient but
almost any other type of loan, including a bank overdraft, is less expensive.
‘Payday loan’ rates are the exception; they are much higher again.
ACTIVITY
A discussion of debt and types of loan with pairs modelling explicit examples
with numbers on a spreadsheet like that in Figure 9.3.1 will reinforce the po-
tential value of slow thinking when it comes to incurring debt.
Nonetheless, debt can be advantageous in some circumstances – we then call

it credit! Sellers sometimes offer very low, even zero, interest rates in order to
clinch a sale, effectively spreading the cost over a period – with car sales often
up to three years. However, this money is not free for the seller, so it must be
built into the asking price.
9.4 Thinking longer term
Borrowing
There are circumstances in which borrowing capital can save a lot of money,
despite Polonius’ splendidly pompous advice in ‘Hamlet’:
Neither a borrower nor a lender be

For loan oft loseth both itself and friend
While borrowing dulls the edge of husbandry.
FIGURE 9.4.1 Change in UK house prices and average wages since 1970.
Source: Data from UK Office for National Statistics [9J].
Buying a house is the most familiar example. Houses in the UK have been
an excellent investment over many decades, with prices rising much faster
than average wages as shown in Figure 9.4.1 – and typically about 3% above
inflation, although with periodic dips that have cost some buyers. Indeed, for
most people with any assets, their house has been their main investment. So
‘getting on the housing ladder’ has rightly been a goal for many young peo-
ple. However, for those with median incomes and no capital, it has proved
impossible in recent years. Rents follow the upward trend, further ‘squeezing’
the standard of living of those who rent.
ACTIVITY
Discuss how rising house prices divide society into ‘haves’ and ‘have nots’?
What might be done to change the situation?
The graph in Figure 9.4.1 reflects other factors, including a link between
house price increases and interest rates. In the early 1990s, rates touched
15% reducing demand for houses and thus prices, while the very low levels
of interest rates in the 2010s have been a spur to house price rises. It may
therefore be a mistake to assume that house prices will continue to generate
such returns in the future, though the primary cause – the excess of demand
over supply caused by the lack of sufficient housing – is not going to disap-
pear any time soon. This is an interesting example of the many factors that
can influence the price of a capital asset –a topic on which students may wish
to study further.
Investment
It is unlikely that many students in a class will have significant sums for a
portfolio of investments but they may be interested in role-playing financial
planning scenarios. There is plenty of data on the performance of various
types of investment over time. This can provide valuable experience in han-
dling data – particularly in graphical form.
The fundamental choice is risk versus reward. Financial advisers com-
plement the values issue with another set of questions to focus on how far
the client wants to take risks in order to increase probable-but-uncertain
returns. They look for consistent, at least semi-quantitative, measures of
their client’s appetite for risk. To this end, they have devised structured se-
quences of questions. Some are general and other questions present specific
situations where the client chooses an option, as in the following illustrative
examples:
• If you had to choose between more job security with a small pay increase
and less job security with a big pay increase, which would you pick?
• Investments can go up or down in value and experts often say you
should be prepared to weather a downturn. By how much could the total
value of all your investments go down before you would begin to feel
uncomfortable?
• Imagine that you are borrowing a large sum of money. It’s not clear which
way interest rates are going to move – up or down. Would you borrow
at a variable interest rate or a fixed interest rate that is 1% more than the
current variable rate?
From the answers to the questions, a client’s risk score is calculated. This is
another example of a mathematical model of an abstract concept (risk appe-
tite) being created and used as a measure from which decisions can be made.
Chapter 12 discusses the topic of creating measures. As with so many instru-
ments used in social sciences, the operational issue is not whether the results
are ‘true’ but whether they are consistent – which is evaluated during the de-
velopment of the instrument – and whether they are useful for their declared
purpose. Here, on the basis of the risk score, supported by the responses to
individual questions, an adviser will recommend different mixtures of invest-
ment offering an appropriate balance of risks versus potential returns. In
general, savings accounts with banks have low risk and low returns, bonds
and unit trusts have more risk and moderate returns, while individual shares
promise high returns at the cost of higher risk. Within each category, some
products will have more risk and potentially more gain than others; for ex-
ample, for shares in different companies, it depends on the kind of business
as well, of course, as unexpected factors like management changes.
ACTIVITY
This ‘investment game’ gives students experience – direct but hazard-free – of

the principles set out above. The app can be run via link [9E] and a screenshot
is given in Figure 9.4.2.
The table and graph show the value of ten company shares over five years.
Click on the table to see information on each company. This data is fictitious,
but loosely based on actual stock market data for the period.
Work in pairs. If you are in a class, don’t give away which shares you choose
but do tell everybody how much money you have made or lost!
• Imagine you have £10 000 to invest and you can choose one company’s
shares to buy in 2017.
• Click on your company and click on the Invest £10 000 button. The app will
show you what happens to your investment.
• Share results with other pairs and discuss what you have learned about
making and losing money.
• Now reinvest by dividing the £10 000 equally across five of the shares
(without knowing how they would perform). Share and discuss the results
and relate them to decisions about risk.
FIGURE 9.4.2 The Investment Game. Can be run online [9E].

This already complex picture is actually greatly simplified. There are many
key attributes of companies that a good fund manager will look for; for
example, a strong market position, unique/protected products and services,
and strong balance sheets. The finance industry has created, and continues to
create, an enormous variety of ‘products’. It makes money using your money.
Finding your own way through all the possibilities is complex; getting profes-
sional guidance is normally wise.
9.5 Other topics

There are far more topics in finance relating to mathematical literacy than
we can do justice to in this chapter. Some of the major omissions are outlined
briefly below.
Basics
We have not attempted to cover important basic facts that are set out in multiple
sources, and courses on financial literacy. For example, students need to learn
to read documents such as bank statements, payslips, invoices, receipts, and
tax documents, and there are many terms that need to be understood precisely.
Insurance
Here you pay a company with (you hope) plenty of capital to recompense you
in the event of a specific loss or unexpected debt. As the company must cover its
administrative costs and make a profit, a useful guiding principle is as follows:
Don’t insure against a loss that you can cover yourself without disrupting
your lifestyle or plans.
So things like insuring for the maintenance of your domestic appliances after the
manufacturer’s guarantee ends – a common type of solicitation – is probably
not a good deal. Indeed, this is the period of lowest risk of breakdown – after
the initial ‘bugs’ have shown up and before wear and tear causes problems.
There are many examples like that. The bathtub curve discussed in Chapter 7
(Figure 7.2.2) is relevant here.
There are types of risk that should be insured. Some are legally required –
notably car insurance against damage or injury to someone else, the ‘third
party’. Some relate to other risks that you cannot afford to cover – notably
loss or substantial damage to any house that you own, or the Stradivarius
violin you are lucky enough to play and, in some countries, medical insur-
ance. The language of insurance can be complex, including terms such as
coverage, premiums, excess, exclusions, and the precious ‘no claims bonus’,
which need to be well understood.
The present value of future income

In finance, ‘discounting’ is the process of determining the current value of a
payment that is to be received in the future. This is an aspect of financial rea-
soning that is rather more subtle conceptually – and interesting mathemati-
cally. It reflects the fact that the money, if received now, could be earning
compounding interest over the period of delay, though the discount rate is
influenced by other market factors than the current interest rates – such as
inflation.
The discounting factor Dn after n years at a constant discount rate of r%
is given by
1
Dn =
(1 + r ) n
This is illustrated in Figure 9.5.1.
Further study
In this chapter we have focused on the big picture, on the fundamental
principles that underlie financial literacy, and how they may be explored in
classrooms. There are a number of professional bodies involved in the field
of financial education, and some of these produce helpful financial guides
that provide an introduction to most of the topics discussed above [9F]. An
example is the Chartered Institute for Securities and Investment.
FIGURE 9.5.1 The present value of $1000 in the future at constant discount rates.
Source: Economics of Climate Change: A Primer – Ch. 3 [9H].
Acknowledgements
Our thanks to Andy Jervis and Joe Fawcett of Chesterton House Financial
Services for their input on this chapter, particularly on investments. If you
need real investment advice, please consult an expert directly!
Figures 9.3.3, 9.3.4, and 9.4.1 include data from the UK Office for N
ational
Statistics [9J], which is public sector information licensed under the Open
Government Licence v3.0. The inflation data before 1949 in these come from
O’Donohughe et al. (2004) – however, Kate Rose Morley’s Historical Inflation
Rates and Price Conversion Calculator at iamkate.com [9D] was helpful when
preparing the section on inflation.
Figure 9.5.1 is taken from Economics of Climate Change: A Primer, public
domain material from the US Congressional Budget Office (2003).
References
Bachrach, B. (2000). Values-based financial planning: The art of creating an inspiring
financial strategy. Aim High Pub.
Dickens, C. (1850). David Copperfield. Project Gutenberg. https://www.gutenberg.
org/ebooks/766
O’Donoghue, J., Golding, L., & Allen, G. (2004). Consumer price inflation since
1750. Office for National Statistics.
Sawatzki, C. (2016). Lessons in financial literacy task design: Authentic, imaginable,
useful. Mathematics Education Research Journal, 29(1), 1–19. https://doi.org/
10.1007/s13394-016-0184-0

purposes:
[9A] Knowing Where You Are spreadsheet

https://ltml.mathlit.org/9A
[9B] Doubling Time spreadsheet
https://ltml.mathlit.org/9B
[9C] Savings Calculator spreadsheet
[9D] Historical UK Inflation Rates and Price Conversion Calculator – Kate
Rose Morley
https://iamkate.com/data/uk-inflation/
[9E] The Investment Game app
https://ltml.mathlit.org/9E
[9F] Sources of Financial Guides

Chartered Institute for Securities and Investment https://www.cisi.org/
[9G] Financial Literacy Analytical and Assessment Framework – PISA/
OECD
https://www.oecd.org/pisa/publications/pisa-2021-assessment-and-
analytical-framework.htm
[9H] The Economics of Climate Change: A Primer – Congressional Budget
Office
https://www.cbo.gov/publication/14387
[9J] Inflation and Price Indices – UK Office for National Statistics
https://www.ons.gov.uk/economy/inflationandpriceindices
10
COMPUTERS IN TEACHING FOR
MATHEMATICAL LITERACY
10.1 Overview
There are many ways in which information technology (IT) can assist the
teaching of mathematical literacy by putting powerful mathematical tools
in students’ hands, making real-world information easily available in the
classroom, and presenting it in rich, interactive forms. Sometimes, this will
also teach useful lessons about the use of such technology in work and per-
sonal life – but the time available, and the rapidly moving target of current
technology, makes the teaching of valid, transferrable workplace IT skills
a challenge, while the actual software tools used in the workplace can be
highly specialised and complex to learn. So here we start by concentrating
on ways IT can be used to help address the particular challenges of teach-
ing for mathematical literacy, expanding on the multiple examples that ap-
pear elsewhere in the book. In Section 10.2, we start by looking at ways
a computer can help in the modelling process, moving in Section 10.3 to
roles in visualisations and simulations, and focusing on the use of video as
a stimulus in Section 10.4. Section 10.5 addresses using realistic data sets,
so central to mathematical literacy, while Section 10.6 is about error detec-
tion. Section 10.7 seeks to bring out the importance of units and orders of
magnitude in technology. We conclude by pointing to some further oppor-
tunities in Section 10.8.
Of course, this chapter can only scratch the surface of the vast set of pos-
sibilities for using IT in teaching that are opening up to schools, and we have
not attempted to address the teaching of coding, computer science, or the
advanced use of software tools.
DOI: 10.4324/9781003303503-11
192 Computers in teaching for mathematical literacy
IT and the need for mathematical skills

Arguably, other subject areas have done more to embrace the use of tech-
nology – such as word processing, making presentations, researching infor-
mation online, creating music or art – while, in mathematics, even 1970s
pocket calculator technology is still sometimes contentious. Perhaps this is
precisely because digital technology is so intertwined with mathematics – and
thus disruptive to the traditional curriculum. A word processor won’t write
a good essay for you (although the qualifier ‘good’ is becoming important –
see below), but the right program can compute the correct answer to an
extremely complicated mathematics problem. The fundamental usefulness
of a word processor doesn’t vanish if you disable the spelling and grammar
checker (which are major aspects of traditional school curriculum). By con-
trast, the fundamental usefulness of a computer algebra system is to compute
answers as required in traditional mathematics problems (including step-by-
step working if you wish) and a spreadsheet that doesn’t do calculations is
not very useful.
As we have seen, technology has changed the aspects of mathematics that
are important, largely obviating the need to perform laborious calculations,
graphing, or algebraic manipulations in most life and work situations. In-
stead, it has made it important to understand the underlying mathematics
and how it models the situation at hand. It has not removed the relevance
of more advanced mathematics to anybody wanting to pursue the subject
into higher education. With vast amounts of data on hand, choosing the
tools to analyse and represent data is essential, as is correctly interpreting
those results – the procedural details of how to perform those analyses,
or construct those representations less so. Any claims that repeated prac-
tice of a procedure will automatically engender better understanding of the
underlying mathematics are not supported by research on learning; rather
the cognitive load of mastering the procedure can overwhelm any deeper
understanding. There are better ways to understand place value notation
and associativity than practising long multiplication and division. As stated
in Hoyles et al. (2007) when reporting on studies of “Techno-Mathematical
Literacies” needed in the modern workplace: “the major skills deficit for
mathematics in workplaces is the understanding of systems, not the ability
to calculate”. Later in this chapter we examine just two cases – spotting
unreasonable results and dealing with ‘new’ units and dimensionality – of
mathematics which has been given new importance by the ubiquity of infor-
mation technology.
At the time of writing, the teaching of languages, arts, humanities, and
social sciences, where assessment revolves around the essay or portfolio of
work, appears to be facing an analogous challenge from widely available
‘artificial intelligence’ software that can write convincing (if uninspired) prose
Computers in teaching for mathematical literacy 193
and poems and create paintings on a given subject. Here too, being able to
use such tools correctly, honestly, and productively, and being able to under-
stand and improve on the results, is likely to become the important life skill,
requiring a change of emphasis in teaching somewhat analogous to those
being discussed here for mathematics.
IT as a teaching tool
It is now common for mathematics classrooms in wealthy countries to have,
at least, an electronic whiteboard with an internet connection – providing
access to a vast bank of resources, some especially designed for education.
Often, these resources are used to deliver fairly conventional course materi-
als – but they have huge potential for introducing realistic content, such as
genuine data sets and images, current news, and video material into a lesson.
For nearly a century, there have been excellent educational films and videos,
but their use has tended to be passive, whereas modern technology allows
the more interactive use of short video snippets, animations, and interactive
simulations that can help students engage with quite complex concepts and
problems.
In the ‘Making and Selling a Magazine’ example lesson below, students
go beyond straight-line graphs and locate the maximum of what looks
quadratic but they use only basic arithmetic. In Chapter 9 it is shown how
students can explore exponential functions, again using simple arithmetic
and iteration. Technology makes it less tedious to explore different cases
and avoids getting rice all over your chessboard. In the past, equations that
cannot be easily solved algebraically had to be avoided. Now nearly any
equation can easily be approximately solved visually by graphing it – in
multiple dimensions if necessary – and zooming in on a root (if there is
one!). Indeed, in industry and science, solving equations by (sophisticated)
‘trial and improvement’ is much more prevalent than solving equations
analytically.
10.2 Computers in the modelling process

Chapter 1 discussed the mathematical modelling process and how it is widely
applicable to many real-world applications of mathematics and described
the cyclical ‘modelling process’ shown in Figure 10.2.1. In a traditional pen-
and-paper mathematics classroom, the technical skills involved in the ‘solve’
step (actually called ‘compute’ in some versions of the modelling cycle) are
often the focus of the lesson at the expense of the other phases – and also the
bottleneck that restricts the complexity of the model, what mathematics can
be used, the size of the data set, and how many iterations of the modelling
cycle can be completed. ‘Solve’ is expensive in terms of the required skills,
FIGURE 10.2.1 The modelling process. See Section 1.3.
the time taken in laborious calculations, and the total cognitive load on the
students. With ubiquitous technology, ‘compute’ is often cheap, instantane-
ous, and can sensibly be a ‘black box’ with its internal workings remaining
unrevealed in the lesson. Technology brings the opportunity to properly fo-
cus on the other three steps of the process.
Modelling with spreadsheets

Aside from the pocket calculator (whether in your pocket or on your wrist,
phone, or computer), possibly the closest thing to a ubiquitous, general-
purpose mathematics tool is the spreadsheet. Spreadsheets first appeared with
Visicalc in the late 1970s and played a significant role in popularising early
personal computers as serious business tools. Unlike older business software,
spreadsheets were highly interactive. Their main original use was building
and exploring “what-if?” financial models. The magazine task below is a
very simple example. Over time spreadsheets have also become widely used
as graphing, charting, statistics, and simple database tools.
In schools, the use of spreadsheets has tended to be dominated by handling
survey data, particularly for collating in tables, summarising with simple
statistics, and graphing. Less attention has been given to the use of cell refer-
ences and formulae to build actual mathematical models. When the ‘Making
and Selling a Magazine’ task in Figure 10.2.2 [10A] was being designed, it
received criticism for being such a basic use of spreadsheets. However, dur-
ing trialling, it emerged that while students were familiar with tabulating
survey results, producing bar and pie charts and summary statistics, the idea
of using formulae to turn that into a “what-if?” model was novel. The video
even shows one pair using a calculator app to calculate values to type into
spreadsheet cells.
The task in the Magazine lesson is to decide on the selling price for a
product, covering the cost of producing the product and maximising profit.
FIGURE 10.2.2 Making and selling a magazine.

Source: Bowland Maths, 2008 [10A].
Will you make more money by selling lots at a low price, or fewer at a high
price? The lesson design assumed that some students would only have mini-
mal spreadsheet experience and so provided a template to ensure they could
concentrate on the formulation aspect of the task and have time for discus-
sion and reflection. A mathematically literate student should eventually be
able to create this sort of spreadsheet themselves, using formulae, creating
well-formatted graphs with sensible scales, and using a few more advanced
tricks such as static references and named cells – all of which requires teach-
ing time in mathematics or digital technology subjects. A video account [10B]
of a lesson built around this task is available, including a useful example of
how a teacher might introduce the scenario.
This task predates the decline of physical printed media, but it works
equally well for the many other small products that students might make and
sell (e.g. hair ties, decorative bag tags, snacks). It could start by asking the
students to think of a product they could make and sell. The mathematics of
supply, demand. and production costs remain the same.
The Magazine task is followed by a suggestion for a data gathering and
analysis activity. Gathering their own data should help students engage
with the subject. However, gathering data can be time-consuming, so it is
important to allocate enough time for the later stages where students dis-
cuss the results, thinking critically about the model and its limitations, and
the sampling. Experience of simple modelling like this can help build some
‘know about’ understanding of how modelling is used to answer serious
problems.
Beyond using formulas and making graphs, a mathematically literate stu-
dent should be able to use at least some of the most common built-in func-
tions, such as sum and average and trendline. The ‘Height and arm span’
lesson outlined in Figure 10.2.3 is intended for students studying the algebra
of linear functions in early secondary school. It begins by using the table
and scatterplot functionalities of spreadsheets, but then uses the ‘black box’
capability to draw a trend line through data. At this early stage in learning al-
gebra, the intention is to focus on the meaning of all four letters in y = mx + c,
to link tabular, graphical, and symbolic representations and to see how func-
tions might link to data and real-world problems. There is no intention for
students to understand theory behind the trend line, for example, using least
squares. Students seem to find the concept of a ‘line of best fit’ sufficiently in-
tuitive. Many variations of this lesson are possible, matching different stages
of students’ learning. For example, instead of using the automated trend line,
students could enter formulae for various lines, and use the spreadsheet to
calculate a simple measure of goodness of fit by summing the absolute differ-
ences between observed and predicted heights. An alternative version, using
a graphics calculator for easy graphing of test lines, is given by Asp et al.
(2004, p. 8).
Height and Arm Span: A lesson on linear functions and data
In this lesson, the whole class investigates the claim that the arm span of a
person is usually equal to their height.
Supply several tape measures (or possibly a measuring app on mobile phones)
and have students work together to quickly measure and records each student’s
height and arm span (the maximum distance from fingertip to fingertip with
outstretched arms).
Enter the data into two columns of a spreadsheet and share to group or
individual computers or graphics calculators. Students create scatterplots and
discuss whether the data supports the claim or not. Then the spreadsheet ‘trend
line’ function is used to display a line of best fit (without formal definition), and
the class can compare it with the claim (y = x). They discuss what the graph of
the trend line shows, and how the trend line can be used for predicting arm span
from height (or vice versa).
When the equation of the line is displayed, the class discusses the meaning of
the numbers in the equation in mathematical and everyday terms. For example,
if the equation is arm span = 1.2 × height – 6, the relationship can be described
as “when a person grows by 10 cm, their arm span will probably increase by
12 cm”. Discussion of the intercept will lead to observations about the range
of validity of the trendline, and aspects of the data set, such as whether it is
representative of the intended population. The automated option of setting the
intercept to zero might be investigated. There are also opportunities to discuss
other aspects of the data (e.g., outliers). The teacher could also supply some
previously collected data from younger/older children or someone very tall.
FIGURE 10.2.3 Height and arm span: a lesson on linear functions and data.
Source: Adapted from Asp et al. (2004).
10.3 Visualisations and simulations

Technology has transformed the way data can be visualised. The summary
statistics (mean, median, mode, correlation coefficient, etc.) typically stud-
ied in school mathematics are still important, but some of their usefulness
harks back to the days when plotting a graph from a large data set was a
time-consuming task. Looking at scatter graphs and other representations of
data should now be a crucial first step in analysis. Whereas it is dangerous to
draw conclusions from data without significance tests, it is just as dangerous
to apply traditional line/curve fitting or summary statistics (all easy-to-use
features of spreadsheets and statistical packages) without some idea of the
distribution of the data. (Does a relationship look reasonably linear? Does a
distribution look like it has a meaningful central tendency?)
Something as simple as a clear animation showing a process or a simula-
tion based on entering one or two parameters can help to reinforce students’
understanding of a concept. The ‘Random variations’ applet in Figure 4.2.4

in the section on ‘Risk’ is an example of a simple simulation used to illustrate
a point. It shows the power of being able to run a simulation many times
to see the effect of random elements. Using simulations is not just to rein-
force mathematical concepts. With simulation as the ‘third pillar of science’
(Weinzierl, 2021) complementing the two long-standing pillars of theory and
experiment, using simulations also helps to develop ‘know about’ insight into
the way in which simulations can be used for solving complex problems.
There is some research (Ridgway et. al., 2007) to show that interactive
visualisations on computers can allow students to engage with multivariate
data more easily than when the information is presented on paper. The task
shown in Figure 10.3.1 presented a simulated data set from a photosynthesis
experiment in which the rate of oxygen production (dependent variable) was
affected by two independent variables (temperature and light intensity). Stu-
dents could put one independent variable on the horizontal axis of a graph
and vary the other interactively using a slider. The actual task asks students
to use the data from the graph to comment on a series of statements from
fictitious students, such as “Jim suggests that they keep the plants as warm as
possible; Amy suggests that they keep the lights on full power – who has the
FIGURE 10.3.1 Oxygen activity from the World Class Arena project (2001). As
discussed in Ridgway et al. (2007) [10C].
best idea?” In the study mentioned, performance on a series of tasks like this
was compared with paper-based graph interpretation tasks which avoided
the use of multivariate data. Despite the richer, more realistic data sets and
the added complexity of multiple variables, students only found the multi-
variate tasks slightly more difficult than the paper ones.
Visualisation tools could be custom-written ‘apps’ or pre-prepared files
for spreadsheets, graphers, or interactive geometry packages. In the reSolve
‘Cornering’ module, a series of lessons investigates the geometry of large
vehicles turning corners, and the implications for the safety of other road
users and for road design. To support this, the materials include a set of
interactive geometry models of increasing complexity – supplied as files for
the free interactive geometry tool GeoGebra – these help students visualise
the situations and explore the effect of various parameters, such as the width
of the vehicle and the length of the wheelbase. These interactive models help
students visualise the situation, and as simulations of the real situation, they
provide predictions about the space vehicles need to turn. All the online live
models and files to download and perhaps modify are accessible from [10D].
The simplest situation is a bicycle or a scooter, and students can do their
own experiments with them to get insight into the geometry of turning.
Figure 10.3.2 illustrates the next stage, where students investigate the turn-
ing of a very long bicycle (a simplified long truck). Here the model primarily
helps visualisation of a simplified case, for which students could reasonably
FIGURE 10.3.2 Bicycle cornering.

Source: From reSolve: Maths by Inquiry (2018) [10D].
FIGURE 10.3.3 Turning into a parking space.

Source: From reSolve: Maths by Inquiry (2018) [10D].
be expected to follow the mathematics. With the interactive model, the em-
phasis can be on exploring the effects of the parameters, rather than repeat-
edly resolving the problem manually.
By the end of the unit, as in Figure 10.3.3, the model has become more
complex, with multiple parameters controlling the dimensions of the (now
four-wheeled) vehicle, as well as the size of parking bay and the turning space
allowed. Although these models use freely available interactive geometry
software, which can be edited, it is not the intention that students construct
the model for themselves. A handful of enthusiastic students might like to
do this, but at this point the students are drawing together all that they have
learned, using the model to make and report the recommendations for the
design of parking spaces.
10.4 Video as a stimulus

A short video or animation can be used to inspire students’ interest and make
the mathematics feel relevant. The use of “drama” in lesson design is dis-
cussed further in Chapter 11 and many of the examples there make use of
stimulus videos. A straightforward mathematics activity using video stimuli
is simply to present a video showing a changing situation, which students are
asked to represent on a graph. Several examples of this are currently avail-
able on the Desmos website [10E] under the heading of ‘Graphing Stories’.
The lesson example “Will It Hit the Hoop?” is taken from the work of Dan
Meyer and others: many others have created versions of this task. The video
shows the start of several basketball shots, and students are asked to predict
if each shot will go through the hoop and sketch the path. On the Desmos
site, this has been built into an interactive activity with graph-sketching and
curve-fitting tools for quadratic functions. Figure 10.4.1 shows a frame from
the video at the top and the manipulable parabola at the bottom. Students
FIGURE 10.4.1 Graphing Stories example: video (top), curve fitting tool (bottom).
Source: From Desmos [10E].
use their best parabolas to make more predictions using the ‘In” and ‘Out’
buttons supplied and then by seeing the full videos of the shots, compare
the by-eye and parabola predictions. This is also an ideal way of practising
plausible estimation so important to mathematical literacy; even the graph-
ing activities generally require choosing a plausible scale.
Students will enjoy making their own videos for activities such as this or
taking still photos of local features to analyse; maybe this bridge is a parab-
ola, maybe a hanging chain or a rainbow is a parabola. However, they will
need some advice about getting suitable images: straight on, level, without
camera distortion. Spreadsheets and dynamic geometry programs (including
Desmos’ tool) can be used to analyse images. Good free software is available
for analysing videos, and some students might have seen this when learn-
ing to play a sport or using sensor kits in science lessons. Pierce and Stacey
(2006, 2011) provide many examples of using videos and images to learn
about geometry, functions, and algebra.
Many of the example activities in this book – where they don’t already
use video or online data – could be enhanced by starting the lesson with,
for example, a current clip from a news site. Videos and animations used in
this way can help address one of the pitfalls of using realistic applications of
mathematics in the classroom: the need for long, high-reading-age texts to
describe the situations.
10.5 Using realistic data sets

For practical reasons, classroom data handling has traditionally been limited
to tiny data sets which can be realistically managed using pencil and paper.
Computers make it possible to work with larger, richer, and realistic data sets,
and – as noted above – interactive visualisation and the freedom from the bur-
den of calculation can allow students to explore more complex relationships
that they would on paper. Many classrooms now have easy access to large
quantities of real-world data online, including resources that have been specifi-
cally curated and developed for educational use. The following two lesson ex-
amples illustrate what can be done with good design and careful development.
Habitat modelling following fire

This lesson, intended for Year 9 and Year 10 school science and mathemat-
ics, is from the Victorian Department of Education Numeracy Across the
Curriculum project [10F]. Ideally it is taught in conjunction with work in
Science or Geography, so teachers with expertise in ecology can contribute.
The lesson illustrates the mathematical literacy principle that the context
needs to be taken seriously and gives a practical insight into how scientists
use mathematical models. It draws on data collected from mallee habitat,
which is widespread over drier parts of southern eastern Australia. Mallee

is dominated by small multi-stemmed trees, sandy soil, and is home to many
animals. Fire is a regular feature of mallee. The data used in this lesson was
collected to study how the habitat responds in the decades after fire.
Scientists surveyed over 500 mallee sites with a known date of the last fire.
They measured habitat features, like tree heights, litter, logs, and grasses and
counted wildlife and plant species. The lesson is based on data that spans up
to 110 years after fire for the variables related to the trees, Triodia (a dense
spiky grass commonly called spinifex) and the ‘legless lizard’ Delma austra-
lis. Some of the data from the scientific publications is presented in a series
of animations created by Ian Lund [10G]. The animation shows the changes
over time. A column graph is used to show the value of each variable (tree
height and cover, stem density, number of hollows, bark, numbers of Triodia
and Delma). All the variables are scaled to a common range of 0–100. As the
animation runs, the maximum value reached up to that time is marked with
a red line on every column.
The lesson begins with teacher-led class discussion, ensuring that students
understand the ecological variables and the data that are involved (e.g. what
live hollows are and why they are important) and the way the animation
presents the data over time. Students then might work in groups to look at
the main trends shown by the animations. Some would be expected (trees
get higher over time), whilst other trends are less easy to see and some are
unexpected (tree cover reaches a maximum and then drops). Once students
are comfortable analysing the seven habitat features over time, they work in
pairs using a second animation which also shows the prevalence of the legless
lizard. Students describe the prevalence of the lizard over the decades after a
fire (it rises, slowly at first, then falls) and they identify the habitat features
which are associated with this. Finally, students make predictions about what
will happen in the short term and long term if there is a fire next summer,
supporting their predictions with evidence and critiquing alternative conclu-
sions from other pairs. Working together enables students to consider multiple
interpretations, and become aware of other observations about the data.
This lesson illustrates a number of features about mathematical literacy.
Understanding many aspects of the context is essential to interpret the data,
and especially to make predictions from it. A teacher knowledgeable in ecol-
ogy is needed to assist students learn why tracking changes in habitat over a
long time after fire is important, what mallee habitat is like, why the selected
variables are important, and how the data set has been assembled. The im-
portance of work like this is heightened because as climate change is likely
to cause more frequent burning and so the time between burns is likely to
reduce – what will be the impact of this? At first glance, the only mathemat-
ics required is to be able to read column graphs, but many mathematical
problem-solving skills and scientific inquiry skills need to be brought into
play to deal with the complexity of the situation. Moreover, there are hidden
mathematical aspects behind the presented data – how is data put onto the
common 0–100 scale, how did the 500 sites provide data for a century, what
sort of sampling was used to get the measurements (e.g. of the number of leg-
less lizards), not all the trees will be of same height so what measure is used,
and how reliable are the measurements of all variables likely to be. Central to
this is the critical thinking that students have to engage in for testing out the
potential patterns that they see, for providing evidence-based arguments to
support their decisions, and for evaluating the arguments of others.
Open data example – literacy rates

Organisations such as the World Bank, UNESCO, Gapminder, and even the
CIA publish large collections of data online that can be used to explore a
wide range of topical social and economic issues – either through those or-
ganisations’ own online visualisation tools or by downloading the raw data
[10H]. There are also large scientific and environmental data sets, often with
real-time updates available, many especially designed for education. The
Australian science agency, CSIRO, for example, has educational data sets on
topics such as the Great Barrier Reef, carbon dioxide concentrations, water
quality in a dam, and inequality of income around Australia.
The World Bank data offers a wealth of demographic and economic data
from around the world along with easy-to-use tools to graph it. The example
below shows literacy rates over time. Data is also available showing the dif-
ference in literacy rates amongst males and females, as well as several statis-
tics about schools and teachers that help to explore this topic. For example,
Figure 10.5.1 shows two graphs of literacy of the whole world population
created using the site. The ‘gender parity index’ is the female literacy rate
divided by the male rate – how is that interpreted?
The site also allows the download of complete, detailed data sets that can
be analysed using a spreadsheet or other tool. Figure 10.5.2 (overleaf) shows
a “deeper dive” into the literacy data, with a scatter graph of male versus
female literacy rates by country or region in 2021. The black line shows what
you would expect if male and female rates were equal. Check that students
know that if a point is on this line, it means the gender parity index is 1, and
that nearly all the data points are below the line because female literacy is
below male literacy nearly everywhere. Which are the exceptional countries,
and what are their characteristics? What might be the relationship between
male and female literacy? It is obvious that a straight line may not be the best
fit. The dotted line shows the quadratic function:
m2
f = , where f = female literacy %, m = male literacy %
100
FIGURE 10.5.1
Graphs of the world literacy rate and gender parity index over
time.
Source: Graph created on the World Bank Open Data site [10H].
FIGURE 10.5.2 Male versus female literacy (%) across countries in 2021.
Source: Created with Data from The World Bank and UNESCO. Spreadsheet available online
[10K].
which is a better fit. But we do not know that this relationship is significant
in any way without more research; indeed, we do not even know if such a
pattern would hold in other years. What is the purpose of fitting a curve to
data like this? Is it just a useful summary of the data or can it be used to make
predictions, or to propose hypotheses to investigate further? Does it give any
clues as to the underlying relationship between male and female literacy?
Certainly, this curved graph highlights the fact that as male literacy increases,
the gap between male and female literacy decreases, but what could be the
cause of that? (Chapter 11 discusses the role of curiosity – part of the produc-
tive disposition component of the context-focused mathematics framework.)
10.6 Spotting ‘computer errors’

Here we move from using IT as a teaching tool to aspects of mathematics that
take on new importance in life and work because of the rise of technology.
As the adage goes, “To err is human – but to really foul things up you need
a computer!” Digital technology has the ability to repeat mistakes millions
of times a second and computers have no common sense or intuition to spot

when something has gone wrong. They follow the instructions supplied re-
gardless. There are plenty of stories of careers and businesses being ended
by an errant decimal point in a computer entry [10L], supporting the well-
known saying in computing – ‘garbage in, garbage out’. That this was not
self-evident to most people was recognised by Charles Babbage (1791–1871)
even before working computers existed:
On two occasions I have been asked [by members of Parliament], ‘Pray,

Mr. Babbage, if you put into the machine wrong figures, will the right an-
swers come out?’ I am not able rightly to apprehend the kind of confusion
of ideas that could provoke such a question.
Charles Babbage (1864). Passages from
the Life of a Philosopher, p. 67
So, given the possibilities of actual software bugs, user errors, or a combination
of both, it is helpful to have some expectations as to what the correct answer
might reasonably be, whenever using technology to perform a calculation.
Estimation skills have long been covered by most math curricula, often
through estimation in the sense of approximate calculation rather than the
broad sense required for mathematical literacy that is set out in Section 3.4. For
example, students might be asked to estimate 199.8 ÷ 49.9, without detailed
calculation. Often, as here, these questions are contrived to have only one rea-
sonable answer, for the convenience of test scoring. This is not a bad problem,
but it only covers one aspect of estimation and is arguably mostly a test of
rounding skills. A more realistic “error spotting” test might be as follows:
Without using a calculator, decide which one of these calculations is wrong:
199.8 ÷ 49.9 = 4.004

1980 × 9.9 = 19800
1240 × 0.1 = 124
1240 ÷ 0.1 = 12400
A completely different – and usually more demanding – type of estimation

task is the “plausible estimation” or “Fermi estimate” where students are
tasked to estimate some quantity with very little of the information required
for an accurate calculation.
In a video from Bowland Maths [10M], teachers tackle the problem of
estimating how many people can stand on a football (soccer) pitch. They
don’t do a very good job, largely as a result of not knowing how big a soccer
pitch is, or how large their chosen unit (the foot) is. They start by assuming
that the pitch is about 22 feet long (this is less than 7 metres), then that it is
about 1/3 as wide as it is long – at which point one comments that the goal is
at least 8 feet wide so that the width is wrong. Then they backtrack and start
reasoning from the size of the goal relative to the height of the player to get
a better estimate. It is the need to develop this ability to spot an implausible
consequence of a mistake or poor assumption – not the calculations involved
– that makes these ‘plausible estimation’ activities valuable in a world where
technology has immense computational power but zero common sense.
The Coin Counting activity [10N] asks students to estimate the value of
a pile of mixed coins being fed into an automatic coin counter. This is a
perfect example of a scenario of where having a rough idea of the expected
answer would help a user to spot any gross failure of the technology. Having
a reasonable expectation of the cost of items in a trolley is one way to guard
against scanning errors, such as scanning something expensive twice. Bow-
land Maths also includes a classroom project You Reckon? [10P] consisting
of a collection of ‘plausible estimation’ tasks.
Beyond estimation, experienced users of any system perform a range of
checks such as looking at max and min values in a set or keeping tabs on the
number of entries with a given property.
10.7 Units and orders of magnitude in technology

Computer technology in the twenty-first century has introduced to everyday
life a slew of new measurements and units. These were once only relevant
to engineers and technicians – now they figure into common consumer deci-
sions. Pick any advert for an IT product or service and there is probably some
use for mathematical literacy.
ACTIVITY
The following was part of an advertisement for broadband internet:
Fast average download speeds up to 66 Mb.

Great for streaming, video calling, and gaming on multiple devices.
• Does it make sense?

(No – 66 Mb is a quantity, not a speed – they mean 66 Mb/s.)
• Does an average speed of 66 Mb/s guarantee “great streaming”?
(No – streaming video or music needs a guaranteed minimum speed.)
• How long would that take to download a 1 TB file at average speed?
(Be careful – the lowercase ‘b’ usually means ‘bit’, the uppercase ‘B’ usually
means byte normally 8 bits, so there’s at least a factor of 8 there.)
The growth of computing power – including memory and storage size – has
increased exponentially over the last 50 years or so. In fact, Gordon Moore
(co-founder of Intel) speculated in 1965 that the number of transistors that
could be fitted on an integrated circuit (then a major limiting factor in the
memory capacities of computers) would double every two years – literally
(really) an exponential growth – a prediction that has proven to be surpris-
ingly accurate, possibly because the industry saw it as a target rather than
a prediction. So, now, when buying a new laptop, you need to know the SI
prefix for 1012 (T or tera). The amount of data generated worldwide in 2022
was predicted to be 97 zettabytes (Statista, 2021): 1 zettabyte = 1021 bytes.
To deal with this growth in large numbers, two new SI prefixes have been
adopted in 2022: ronna (R) is 1027 and quetta (Q) is 1030 – although quet-
tabyte memory sticks are unlikely to be showing up in the shops just yet.
The much wider range of numbers and units now required were identified
in Section 3.3 as one way in which mathematics curricula must adapt for
mathematical literacy.
Image sizes and resolutions

Computer imaging – now part of every mobile phone – brings more issues of
scaling and dimensionality into the scope of consumer decisions. Converting
between the measures used in advertisements and the details you need to
make decisions can be tricky, for example:
ACTIVITY
Your old phone has an “8 Megapixel” camera. You look at your photos on a
“Full HD” TV which has a vertical resolution of 1,080 pixels and an aspect ratio
of 16:9. If you want to see more of the detail in your pictures, would you
a buy a new “4k UHD” TV which has a horizontal resolution of 3,840 pixels.
(Yes, beware, between “HD” and “4k”, they changed the definition of TV
resolution from vertical pixels to horizontal pixels!) or
b buy a new phone which has a “12 Megapixel” camera.
Whatever – you want a new TV anyway! You’re looking at one with a 120 cm
diagonal screen and a “16:9 aspect ratio”. The rim around the screen is pretty
small, less than 1 cm.
Will it fit in an alcove 110 cm wide? Maybe create a spreadsheet (using
formulae) so that you can work this out for other TVs you’re considering.
A mathematical literacy investigation of image resolution could be ex-

tended to include factors like how the amount of storage (or internet band-
width) varies for different image resolutions.
10.8 Further opportunities

This chapter has only scratched the surface of the potential uses of computer
technology in the classroom, picking out a few cases particularly relevant to
mathematical literacy. Coding is another area that offers potential. As early
as the 1980s, students were learning to code with LOGO (Papert, 1980) and
it was immediately evident that this opened up many new ways of developing
mathematical concepts and also to developing ‘know about’ appreciation of
how computers work. There are now multiple child-friendly versions availa-
ble, most notably Scratch [10Q] where students can program with movement,
music, and colours. Many secondary school science departments will also
have sets of digital sensors that can be used to record attributes such as posi-
tion and speed, light intensity, pH, temperature, dissolved oxygen, and many
others. The kits are also equipped with video and data analysis tools, which
support students bringing their understanding of mathematics to enrich ex-
plorations of many real-world situations. Students could, for example, make
their own videos and graphs to analyse situations like those in the basketball
example in Section 3. Digital sensors used to gather and analyse data are ideal
to support interdisciplinary work across the health and science subjects.
Unsurprisingly, because the rise of information technology is a major fac-
tor in raising the demand for mathematical literacy, more examples can be
found in the other chapters of this book. For example:
• Chapter 2: Mechanical linkages and deductive geometry lessons (Figure 2.4.2)

• Chapter 4: Using a simple app to illustrate the effect of random variations
in the frequency of low-probability events (Figure 4.2.4)
• Chapter 4: Epidemic spreadsheet (Figure 4.4.1)
• Chapter 6: Creating Gannt charts, using graphing software for linear pro-
gramming (Figures 6.3.1 and 6.3.2)
• Chapter 8: Finding and presenting demographic data
• Chapter 9: Financial planning using spreadsheets (Figure 9.1.1), Modelling
interest with spreadsheets (Figure 9.3.1), ‘The investment game’ (Figure 9.4.2)
Acknowledgements
Figure 10.2.2 from Bowland Maths appears courtesy of the Bowland Maths
maintainers. Figure 10.3.1 appears in Ridgway et al. (2007) which is avail-
able under a Creative Commons Attribution licence – the original task was
designed by Jim Ridgway, Daniel Pead, and others as part of the World Class
Tests project.
Figures 10.3.2 and 10.3.3 appear courtesy of reSolve: Maths by Inquiry

[10D] – reSolve materials are available under the Creative Commons
Attribution-NonCommercial-Sharealike 4.0 International (CC BY-NC-SA 4.0)
licence.
Figure 10.4.1 appears courtesy of Dan Meyer at desmos.com [10E].
Figure 10.5.1 was captured from the World Bank website using data from
UNESCO as cited in image. Figure 10.5.2 includes World Bank and UNESCO
data licensed under the Creative Commons Attribution License 4.0.
References
Asp, G., Dowsey, J., Stacey, K., & Tynan, D. (2004). Graphic algebra: Explorations
with a graphic calculator. Key Curriculum Press.
Babbage, C. (1864). Passages from the life of a philosopher. Longman and Co.
Hoyles, C., Noss, R., Kent, P., Bakker, A., & Bhinder, C. (2007). Teaching and
Learning Research Briefing Number 27. https://discovery.ucl.ac.uk/id/eprint/
1515625/
Papert, S. (1980). Mindstorms: Children, machines and powerful ideas. Basic Books.
Pierce, R., & Stacey, K. (2006). Enhancing the image of mathematics by association
with simple pleasures from real world contexts. Zentralblatt fur Didaktik der
Mathematik, 38(3), 214–225.
Pierce, R., & Stacey, K. (2011). Using dynamic geometry to bring the real world into
the classroom. In L. Bu & R. Schoen (Eds.), Model-centered learning: Pathways
to mathematical understanding using GeoGebra (pp. 41–55). Sense Publishers.
https://link.springer.com/book/10.1007/978-94-6091-618-2
Ridgway, J., Nicholson, J., & McCusker, S. (2007). Reasoning with multivariate evi-
dence. International Electronic Journal of Mathematics Education, 2(3), 245–269.
https://doi.org/10.29333/iejme/212
Statista (2021). Volume of data/information created, captured, copied, and consumed
worldwide from 2010 to 2020, with forecasts from 2021 to 2025. Retrieved 12 July
2023 from https://www.statista.com/statistics/871513/worldwide-data-created/
Weinzierl, T. (2021). The pillars of science. In Principles of parallel scientific
computing: Undergraduate topics in computer science. Springer. https://doi.org/
10.1007/978-3-030-76194-3_1

purposes:
[10A] Making and Selling a Magazine – Bowland Maths

https://www.bowlandmaths.org.uk/materials/pd/online/pd_04/pdf/
pd_04_handout_4.pdf
[10B] Video of Spreadsheet Lesson – Bowland Maths

https://www.bowlandmaths.org.uk/materials/pd/online/pd_04/
pd_04_follow.html – Activity 3
[10C] Oxygen activity designed by the Shell Centre for the World Class
Arena project
[10D] Bicycle Cornering – reSolve: Maths by Inquiry
https://www.resolve.edu.au/mathematical-modelling-cornering?lesson=
1671
[10E] Will It Hit the Hoop? – desmos.com
https://teacher.desmos.com/activitybuilder/custom/56e0b6af013382
2106a0bed1?collections=5e827a6e58f1e36e4d220ef8
[10F] Habitat Modelling Following Fire – Victorian Department of
Education
https://www.education.vic.gov.au/school/teachers/teachingresources/
discipline/maths/Pages/numeracy-for-all-learners.aspx
[10G] 100 Years of Habitat Change: An Animated Fire Ecology – Ian Lunt
Ecology
Direct link: https://ianluntecology.com/2014/05/25/animated-fire-
ecology/
Video link: https://www.youtube.com/watch?v=30j1tPc-m1s
[10H] Sources of demographic data – various
https://data.worldbank.org/
http://uis.unesco.org/
https://www.gapminder.org/resources/
https://www.cia.gov/the-world-factbook/
[10J] Educational Datasets – CSIRO (Australia’s national science agency)
https://www.csiro.au/en/education/resources/educational-datasets
[10K] Graph of Literacy Rates
https:/ltml.mathlit.org/10K
[10L] Fat Finger Error – Wikipedia
https://en.wikipedia.org/wiki/Fat-finger_error
[10M] Teachers Trying a Plausible Estimation Problem – Bowland Maths
(2008)
https://www.bowlandmaths.org/materials/pd/online/pd_02/pd_02_
intro_01_footy.html
[10N] Coin Counting, by Dan Meyer – 101 questions
https://www.101qs.com/3199
[10P] You Reckon – Bowland Maths (2008)
https://www.bowlandmaths.org/projects/you_reckon.html
[10Q] Scratch Coding Language & Community – Scratch Foundation
https://scratch.mit.edu/
11
THE IMPORTANCE OF CURIOSITY
What are the attitudes and skills that lead a person to ask themselves the
kinds of questions that we have explored in this book? What makes a pro-
ductive disposition towards being mathematical literate? At the heart of this
attribute is curiosity. The generic question is:
What is going on here?
Unless and until a person asks this question, the thinking that is mathemati-
cal literacy, beyond the purely functional, will not begin. Of course, the same
question is at the heart of doing, rather than just learning, mathematics itself.
In Chapters 4–9 we raised such questions across six domains of real-world
importance. We now want to explore in a little more depth the nature of
curiosity and how it may be developed – by individuals and in the classroom.
In Section 11.1 we sketch the history of thinking about curiosity in philoso-
phy and psychology, moving on to education and mathematical literacy in
Section 11.2. Section 11.3 briefly reviews the range of things that stimulate
curiosity, moving on in Section 11.4 to examples of lesson design strategies
for developing curiosity in the classroom. Section 11.5 looks at what we can
learn from a quite different approach that comes from art and design.
In terms of the context-focused mathematics framework of Chapter 1,
curiosity is a central, but often overlooked, part of the component ‘dem-
onstrating a productive disposition’. A person has a productive disposition
for mathematical literacy when they expect that using mathematics will be
helpful and informative, and have confidence that they will be able to use
their knowledge to get a useful result. Productive disposition also requires a
willingness to persist through obstacles, to be prepared to think carefully and
DOI: 10.4324/9781003303503-12
214 The importance of curiosity
work flexibly exercising initiative. Curiosity about phenomena is an essential

precursor of this kind of thinking.
11.1 What is curiosity?

Everyone is curious from time to time, wondering about something that hap-
pens, even if not actually formulating questions about it and seeking answers.
Some people seem naturally more curious than others. In this section we
explore the nature of curiosity, and how it fits into the broader domain of
‘information seeking’.
Curiosity has long attracted the active interest (curiosity?) of philosophers,
psychologists, neuroscientists, and other thinkers. William James (1899, re-
published 1983) called curiosity “the impulse towards better cognition”.
Daniel Berlyne (1954), an important figure in the twentieth-century study of
curiosity, distinguished perceptual curiosity from epistemic curiosity. Percep-
tual curiosity is seeking out novel stimuli, the primary driver of exploratory
behaviour. It is found in non-human animals and human infants, and is one
driving force of human adults’ exploration. Epistemic curiosity seeks to ob-
tain access to information capable of dispelling uncertainties of the moment,
but also to acquire knowledge. Other perspectives abound. Yet the Kidd and
Hayden (2015, p. 449) in their review of the history concludes: “Despite its
pervasiveness, we lack even the most basic integrative theory of the basis,
mechanisms, and purpose of curiosity”. However, the most popular theory
about the function of curiosity is to motivate learning – hence this chapter.
We adopt a heuristic approach, leaning on results of research where they
seem directly relevant to curiosity’s role in learning and teaching for mathe
matical literacy.
Some people have found it useful to distinguish different kinds of curios-
ity, related to the information being sought. Information gap filling or fact
seeking is often routine, involving little curiosity, though they are important
in resolving ambiguity, dissonance, or even suspense. The curiosity that un-
derpins mathematical literacy is usually focused on looking for patterns and
seeking explanations.
It is encouraging that research on young children confirms every parent’s
experience that curiosity is innate in all of us. Infants explore the world
around them from the very beginning, making their own decisions, insofar
as they can, on what to explore. They are curious learners whose cognitive
development involves imposing structure on the environment they explore.
Piaget, a pioneer of such studies, called young children “little scientists”.
More than half a century of research in both lab and more realistic settings
has developed a rich picture of the curiosity-driven investigative processes
involved in early-years cognitive development. This research has brought
about a broad consensus that infants’ information selection and subsequent
The importance of curiosity 215
learning in empirical tasks are influenced by their current thinking, the stim-
uli in the learning environment, and discrepancies between the two – that
is, the element of surprise (for a review, see Mather, 2013). They are asking:
“What is going on here?”
As children get older, their curiosity naturally moves with the things that
most interest them in their everyday life. For some it may be a particular
sport – news about the teams, the players, the results – or another hobby,
including online games, or just things that occur in their family or their so-
cial life, extended greatly through the world of social media. This continues
for all of us into our adult lives. But some people show more curiosity than
others.
Some educational designers seek to transfer this curiosity, interest, and
enthusiasm to motivate learning in school. This Philosophers’ Stone of

motivation has proven elusive, at least in terms of penetrating mainstream
curricula – but work continues in the ways described below.
11.2 Curiosity and education

It has often been asserted by various reformers that formal schooling drives
their innate curiosity out of many young children. It is true that traditional
curricula, particularly in mathematics and science, focus on having students
learn established knowledge – selected, demonstrated, explained, and illus-
trated by the teacher. Teacher and student questioning are focused on the
understanding of these explanations, and how to perform skills to answer
imitative practice exercises.
It is equally true that this approach has long been challenged, often under
the banner of social constructivism. Pioneers such as Dewey (1916) described
learning as fundamentally an exploratory process built on experience in
learning. Gattegno (1970) saw the priority to be learning rather than teach-
ing (reflected in the TRU framework set out in Chapter 3). He emphasised
that learning takes place in stages – from ‘awareness’ that there is something
to explore through a process of deepening exploration until that aspect no
longer needs attention. In the UK this inspired the formation in the 1950s of
the Association of Teachers of Mathematics, which promotes an approach
to learning mathematics that is based on investigation of a wide variety of
mathematical ‘microworlds’, guided by the teacher but with as much student
autonomy as the particular young person can handle successfully. Leading
associations of mathematics teachers around the world promote a similar
philosophy and support teachers to implement it. As we saw in Chapter 10,
information technology has a huge potential for stimulating and sustaining
curiosity through investigation, one emphasised by Seymour Papert (1980)
who created the Logo programming language for that purpose, with Turtle
Geometry as a core example.
Curiosity is central to investigation: the process of asking what is going on

in each problem situation. It involves the student turning a broad question
into a sequence of ever more specific questions related to the problem situa-
tion, while monitoring their own progress (or lack of it) to see if a change of
approach might be more productive (see Schoenfeld, 1985). Most coherent
attempts at improving mathematics education have seen posing such ques-
tions, as a key component of doing and learning mathematics. The Art of
Problem Posing (Brown & Walter, 2004) explores this issue within math-
ematics in readable depth.
This has been carried forward in the recognition, since the 1980s, of the im-
portance of students tackling non-routine problems, in which curiosity is cen-
tral, as part of their mathematics education. Building on the introspections of
Polya (1945), an empirical tradition of research and development in ‘problem-
solving’ has led to further insights and to well-engineered teaching materials –
see, for example, Mathematical Problem Solving (Schoenfeld, 1985), Strategies
for Problem Solving (Stacey & Groves, 1985, p. 2010), Thinking Mathe-
matically (Mason et al., 2010), Problems with Patterns and Numbers [11A],
Classroom Challenges [11B], Bowland Maths [11C] and reSolve [11D]. Such
materials enable all teachers who are working in supportive environments to
build mathematical problem-solving into their enacted curriculum.
Curiosity in mathematical literacy

Investigation, and thus curiosity, is at the core of mathematical literacy.
Curiosity about real-world phenomena is an essential precursor to seeing
how mathematics can help one understand them. How does an investiga-
tive approach to mathematics teaching generalise to situations in the real
world? The problem situations, tasks, and activities in previous chapters are
designed to demonstrate this – developing mathematical literacy through
‘question posing’ that builds towards ‘insight seeking’.
In the real world (as in genuine investigation into pure mathematics),
problems rarely arise as neatly posed questions for you to respond to; you
yourself have to recognise a problem situation and generate specific questions
as a starting point for further investigation. The stimulus is often a more or
less open problem situation provoking you to ask: “What is going on here?”
For example, Table Tennis Tournament (first presented in Figure 2.1.2 and
discussed more fully in Section 6.1) asked the open questions:
Three of us have agreed to organise a table tennis tournament for our club.
How should we set about it? What information will we need?
A group of students who are used to exploring open problems will iden-
tify relevant issues and associated variables, listing more specific questions,
as discussed in Section 6.1, some of which are basic (how many tables are
available) and others involve significant mathematical thinking. For example,
considering whether to have a knock-out tournament, or a ‘round robin’
league where every player plays every other, leads to the question: how many
games does each format require for a given number of players?
The answers are (n − 1) games in a knock-out versus n(n − 1) / 2 for n play-
ers in a round robin league. This shows that a league format requires either
only a small number of players or a long time, which explains the choice
of knock-out tournaments for professional tennis, and for ancillary tourna-
ments in professional football where the typical league takes most of a year.
In Chapter 6 we looked in more detail at the demands and challenges this
task presents to students and teachers.
11.3 What stimulates curiosity?

In this book we have chosen situations that can excite curiosity in classrooms.
Risk (Chapter 4) is an area that features in the press and other media with sto-
ries that may or may not warrant the concern implied. On planning (Chapter 6),
we all know that we should probably be more systematic about it. We recognize
that what we are told in politics and commerce is often misleading – and no one
likes ‘to be made a fool of’ (Chapter 7). Much the same applies to the ultimately
quantitative domain of money (Chapter 9). We hope the specific cases we have
looked at in each of these areas will excite curiosity. But, as with so much in
education, the real measure of success is how students and the adults they be-
come can apply the elements they have learned to new situations. So what are
the elements of curiosity and how can they be developed in classrooms?
We have noted that curiosity is a natural attribute of human beings from
their earliest years. Nonetheless, people vary greatly in how far they want to
ask questions and explore issues or simply accept the world as it is, something
we all have to do to a great extent – you can’t investigate everything. Here we
ask what are the attributes that tend to make people curious. In Section 11.4
we discuss how these attributes can be developed, in the classroom and outside.
Surprise
Unexpected events are a common cause of curiosity. “What was that?” and
“Look at that!” are natural immediate responses to surprising happenings
and objects. Social media are full of videos of things– some genuine, some
faked – that stimulate curiosity, and attract ‘likes’. But surprise is not always
turned into active investigation. And indeed, becoming surprised relies on
some knowledge of what is normal – not surprising. Nonetheless, deliberately
trying to generate surprises in the classroom is a powerful lesson design tactic.
If surprises arise in the course of thinking things through, so much the better.
ACTIVITY
Small groups exchange examples of things related to mathematical literacy that

have surprised them – and whether they were, or seemed, worth investigating.
Concern
Though extreme anxiety can cause paralysis, when a concerning or threaten-
ing situation arises, most people are moved to look into it – to begin to ask
questions. If your landlord moves to raise your rent, you seek information on
what safeguards there may be in your city. News reports about climate change
raise obvious questions: how much global warming should we expect, when
and with what effects, and what can we do about it? The ‘risky’ situations we
discussed in Chapter 4 are examples where concern inspires curiosity.
The next step may be data-driven and/or theory driven. First, there arise
issues of data reliability. Should I believe the climate scientists or the oil com-
panies who assure the public ‘it is all under control’? The second kind of
follow-up is to try to think through the worrying phenomenon in a scientific
way. On global warming, how can an increase in average temperature of just
2 degrees Celsius, which one would hardly notice on a summer day, have
such profound effects? This may lead to a search for scientific explanations of
the mechanism(s), or broader evidence that past predictions are proving cor-
rect, of the kind in Chapter 5.
ACTIVITY
Small groups look for real or potential sources of anxiety and discuss what
steps they might, or might not, take as follow-up – noting the questions they
would ask.
Opportunity
A more positive stimulus for curiosity arises when new opportunities are
perceived. “Wanted, enthusiastic young people for engineering apprentice-
ships”. “Expedition to the Amazon seeking skilled canoeist”. This kind of
message aims to stimulate the curiosity of good candidates for the task.
But the most obvious area of opportunity is money. We are bombarded
with ‘irresistible offers’ – in print, on television, on social media, and even on
our phones – hawking diverse ways to make money. Those who launch such
enticements aim to stimulate curiosity but, often, to avoid closer investigation.

Many sellers are highly skilled at the latter, even – or perhaps particularly – for
purchases of the worst kind, from substandard products to scams, harmful or
illicit goods. Mathematical literacy, with its emphasis on both critical thinking
and mathematical tools, aims to encourage and help people look more closely.
Basic advice like “If it seems too good to be true, it probably is” is a start but
deeper investigation as we discussed in Chapter 9 is needed to evaluate the
risk–benefit balance that underlies good decision-making.
ACTIVITY
Ask small groups to discuss how they react to approaches that purport to offer
money-making opportunities, including the aspects that help them decide to
move forward, or reject.
Success
An activity that has gone well leads one to persist, to look further into it, or to
seek similar experiences. Among well-known indicators of positive attitudes
and productive disposition is “persists in the activity after the allotted time”.
Every teacher or professional development leader remembers the times when
“they kept going after the bell went” or “the questions and comments kept
coming so we nearly missed lunch”.
ACTIVITY
Ask for students who are striving for excellence in an activity to say something
about the ways in which they motivate themselves to persist.
ACTIVITY
Stephen Anderson said “Surprise can be a very minor change that adds flavour
and variety to an otherwise routine experience”. Is that true? For you?
Knowledge
Beyond any immediate surprise, you are unlikely to be curious about some-
thing that you know nothing about. Loewenstein (1994) suggested that a
small amount of information serves as a priming dose, which greatly increases
curiosity. Consumption of information is rewarding but, eventually, when

enough information is consumed, satiation occurs and later information
serves to reduce further curiosity.
More broadly, a good ‘seed’ for engendering curiosity, and subsequently
mathematical literacy, might be one of an individual’s current interests,
about which they already have some ‘primer’ knowledge and a motivation to
understand more, and which has some authentic mathematical content that
can be ‘surfaced’. For teenagers such subjects are likely to include social me-
dia algorithms, social media dynamics (“likes” generation, dynamics of viral
posts), fashion cycles, dynamics of communication and relationships, climate
change, broad social concerns to build a better world, e-sports, and even the
mathematics of exams.
Curiosity within information seeking

How does curiosity relate to information seeking in the broad sense? Are
they the same? This question has been central to all theoretical attempts de-
scribe and classify curiosity – with the absence of any agreed model. We
take the view that it is most useful to see curiosity as a form of information
seeking where an element of novelty is involved. However, persisting in an
information-seeking activity, even after the minimum information required
to satisfy the immediate need has been found, indicates deeper curiosity. For
example, in areas we know a lot about, professionally or informally, we are
likely to want ‘to keep up to date’. Some of this is routine information seek-
ing but going further involves curiosity.
What the different kinds of stimuli described above have in common is that
they are based in emotion. Curiosity is an initial response that may lead to an
intellectual one – fast thinking perhaps leading to the slow thinking that math-
ematical literacy, mathematics, and all intellectual activity involves. In Sec-
tion 11.4 we discuss the pedagogical challenges involved – both in stimulating
curiosity in students and in encouraging the transition to analytical thinking.
11.4 Developing curiosity in the classroom

We noted in Section 11.1 that the most popular theory about the function of
curiosity is to motivate learning. From a teaching perspective, there are two
rather different challenges:
• How can we support students who are interested in understanding what

is going on?
• How can we stimulate the curiosity of students who see learning as no
more than understanding (or even simply remembering) the things they
have been told?
Though there is clearly an overlap, we will distinguish these as feed-

ing curiosity versus developing curiosity. Over recent decades, a lot of
progress has been made in answering the first of these questions. The sec-
ond challenge remains work in progress. Some lesson designers tend to
assume that all students are naturally curious and investigative which, to
say the least, is empirically ‘not proven’; teachers are less likely to make
that mistake.
Feeding curiosity
Through a mixture of research and creative design over many years, lesson
activities have been developed that support investigation of phenomena in
the world beyond the classroom, as well as within mathematics. Teaching the
modelling of everyday situations – an underlying theme of this book – has
been going on at least since the 1960s when one of us (Burkhardt) introduced
experimental workshops built around the following kind of questions:
• If you are crossing between buildings in heavy rain, do you get wetter
walking or running?
• If you plan to buy a used car, what age of car should you buy and when
should you sell it to minimise the cost?
• What ways are there in which climbing a ladder against a wall can become
unstable and which are most dangerous? This example has been discussed
in Section 6.4 to illustrate the climate discussion of tipping points.
Developing curiosity – drama in lesson design

In their article on design, Burkhardt and Pead (2020) suggested that “Every
classroom is a theatre. Use that fact in lesson design – but make the teacher
the director not the star” [11E]. This was inspired by the work of Dan
Meyer (see below) and Malcolm Swan. Unlike some other subjects, math-
ematics lessons rarely have dramatic tension, a story, or a surprise. Usually,
in theatre terms, the teacher is both director and star. The students are the
audience. The play begins with a long soliloquy, then it’s over! Students just
imitate many times over what they have been shown. The drama design
approach can increase student motivation and problem-solving in a broad
sense.
As we have tried to show throughout the book, the task-based lessons
inherent in mathematical literacy can have a dramatic or topical story in-
herent in the problem situation. (A fantasy context can also be contrived
to introduce more abstract pure mathematical problems.) The key strat-
egy is to engage the students by introducing an element of storytelling or
role-play.
The widely admired designer Dan Meyer devised an explicit structure for
whole-group task-based lessons with the title Three Act Math [11F]. The
dramatic structure is based on a standard model for writing stories:
• Act One introduces the central conflict, an engaging and perplexing one
(i.e. designed to excite curiosity).
• Act Two in which the protagonist (student) overcomes obstacles, looks for
resources, and develops new tools.
• Act Three is a solution-discussion and solution-revealing phase that re-
solves the conflict and sets up a sequel (extension problem).
In the examples he developed, Act One is often a somewhat perplexing

video setting the scene plus a question or two, made explicit by the teacher
or left to emerge from discussion. For example, two ‘back of the envelope’
estimation tasks from Meyer’s 101 questions website [11F] are Cola Pool and
Coin Counting.
‘Cola Pool’ shows the last bottle of the soft drink being poured into a
cylindrical backyard pool and asks: How many bottles of Cola did they
buy to fill that pool? The first question is What information is needed to
answer the question? The information is then provided and the calcula-
tions proceed through to discussion.
‘Coin Counting’ is more demanding – the video shows a pile of coins of
various values. The questions are How much cash is that? How many coins
are there? Here deciding the necessary information and how to find it is
the core challenge.
The three-act lesson approach has spread widely in North America. Ontario,
for example, which has a fine tradition in mathematics education, presents
the three-act approach a little differently, with the goal to spark curiosity
and fuel sense-making. They offer some interesting real-world tasks [11G] as
does the Bowland Maths project in the UK [11C]. An example from Bowland
Maths is “Reducing Road Accidents” (see Section 3.2) in which students are
given a budget and make road safety recommendations to the town council.
Experience shows that lessons with a dramatic structure like these seem to
engage all students, not just the already-curious (TRU dimension 3).
The use of a story element that encourages curiosity can be found, to a
greater or lesser extent, in the design of lessons on standard topics in math-
ematics. Leslie Dietiker (2015) and colleagues illustrated this with a detailed
comparative analysis of the lessons from four different textbook series – here
on the minimum information needed to establish that two triangles are con-
gruent. The four lessons differed substantially in the time for which all the
various combinations of angles and sides were left open as possibilities. For
example, a lesson starting with open questions like “If all three angles are
the same, must the triangles be congruent?” explores a wider space than only
showing that the three standard sets of conditions (SSS, SAS, ASA) are suf-
ficient. In summary, leaving more questions open for longer is a method of
encouraging both curiosity and student agency and autonomy (TRU dimen-
sion 4).
11.5 Further thoughts

In this book, as in mathematics education more widely, we have adopted a
largely scientific and utilitarian approach: how can mathematics help peo-
ple navigate life and its challenges more effectively? A different approach to
education that emphasises the value of curiosity has been suggested by Elliot
Eisner (2002), a pioneer in arts education. In it more importance is placed
on exploration than on discovery, more value is assigned to surprise than
to control, more attention is devoted to what is distinctive than to what is
standard. He further declares his preference for aesthetic ways of knowing
and learning, pointing out that artists develop and achieve their ends through
exploiting emerging and unexpected opportunities along the way. He rejects
the common position, summarised as “science is useful; the arts ornamen-
tal”. His vision of education as an art inspires new questions that serve to
address some of its problems: “How can the pursuit of surprise be promoted
in a classroom? What kind of classroom culture is needed?”
What might this mean for mathematics and for mathematical literacy?
Mathematics has always had a strong aesthetic element. Concise ‘elegant’
proofs are more pleasing than ‘proofs by exhaustion’ that consider every case
separately. Mathematical patterns and structures have immediate appeal for
some people. Dietiker (2015) explores the new visions, values, and practices
that taking aesthetics seriously might open up in mathematics education, in-
forming a reimagining of mathematical experiences in classrooms. She sees
the value of mathematics lessons designed as mathematical stories of explora-
tion, extending this approach to the teaching of regular concepts. This echoes
the concept of dramatic design in Section 11.4, and the mathematicalliteracy
examples there and throughout the book, but with an explicit emphasis on
the aesthetic pleasures and the curiosity that drives the search for insights
involved.
In short, the processes of mathematical literacy should be enjoyable as
well as informative.
Acknowledgement
We are grateful to Leslie Dietiker for informative conversations.
References
Berlyne, D. E. (1954). A theory of human curiosity. British Journal of Psychology,
45(3), 180–191.
Brown, S. I., & Walter, M. I. (2004). The art of problem posing (3rd ed.). Taylor and
Francis eBook. https://doi.org/10.4324/9781410611833
Burkhardt, H., & Pead, D. (2020). 30 Design strategies and tactics from 40 years of
investigation. Educational Designer, 4(13).
Dewey, J. (1916). Democracy and education: An introduction to the philosophy of
education. The Free Press.
Dietiker, L. (2015). Mathematical story: A metaphor for mathematics curriculum.
Educational Studies in Mathematics, 90, 285–302.
Eisner, E. W. (2002). What can education learn from the arts about the practice of
education. Journal of Curriculum and Supervision, 18(1), 4–16.
Gattegno, C. (1970). What we owe children: The subordination of teaching to learn-
ing. Outerbridge and Diensfrey.
James, W. (1899). Talks to Teachers on Psychology, and to Students on some of Life’s
Ideals. Science, 9(235), 90–910.
James, W. (1983). Talks to teachers on psychology: And to students on some of life’s
ideals (Vol. 12). Harvard University Press.
Kidd, C., & Hayden, B. Y. (2015). The psychology and neuroscience of curiosity.
Neuron, 88(3), 449–460.
Loewenstein, G. (1994). The psychology of curiosity: A review and reinterpretation.
Psychological Bulletin, 116(1), 75.
Mason, J., Burton, L., & Stacey, K. (2010). Thinking mathematically (2nd ed.).
Pearson.
Mather, E. (2013). Novelty, attention, and challenges for developmental psychology.
Frontiers in Psychology, 4, 491.
Papert, S. (1980). Mindstorms: Children, machines and powerful ideas. Basic Books.
Polya, G. (1945). How to solve it. Princeton University Press.
Schoenfeld, A. H. (1985). Mathematical problem solving. Academic Press.
Stacey, K., & Groves, S. (1985). Strategies for problem solving: Lesson plans for
developing mathematical thinking. Latitude Publications. Second edition 2006.
Swan, M., Pitts, J., Fraser, R., & Burkhardt, H., with the Shell Centre Team. (1984).
Problems with patterns and numbers. Joint Matriculation Board and Shell Centre
for Mathematical Education [11A].

purposes:
[11A] Problems with Patterns and Numbers – Swan et al. (1984)

https://www.mathshell.com/materials.php?item=ppn&series=tss
[11B] Classroom Challenges – Mathematics Assessment Project

https://www.map.mathshell.org/
[11C] Bowland Maths
https://www.bowlandmaths.org/
Reducing Road Accidents: https://www.bowlandmaths.org/projects/
reducing_road_accidents.html
[11D] reSolve – Australian Academy of Science
https://www.resolve.edu.au/
[11E] 30 Design Strategies and Tactics from 40 Years of Investigation –
Burkhard, Pead (2020)
https://www.educationaldesigner.org/ed/volume4/issue13/
article53/#link12
[11F] The Three Acts of a Mathematical Story – Dan Meyer
https://blog.mrmeyer.com/2011/the-three-acts-of-a-mathematical-
story/
Cola Pool: https:// www.101qs.com/3896
Coin Counting: https://www.101qs.com/3199
[11G] Problem-based Math Lessons & Units – Make Math Moments
https://learn.makemathmoments.com/tasks/
12
DESIGNING MEASURES
What is a measure? A measure assigns a number to a characteristic of an ob-

ject or event. Familiar measures include length and weight, time, and money.
A measure of ‘intelligence’ may simply be the assignment of a number to a
set of responses to a test.
Measures are important. Peter Drucker’s (1995) assertion about macro-
economics is as follows:
If you can’t measure it, you can’t manage it.
Drucker’s assertion surely has limitations. How many of the decisions you
make in your personal life depend mainly on quantitative measures? But we
have seen how important measures, with the choice of variables that they
include, are in understanding real-world phenomena and in making large-scale
decisions. We have seen a lot of measures in this book: the probabilities of
various events in Chapter 4 measuring risk, the atmospheric concentration
of CO2 and its relation to global temperature (Chapter 5), the length of the
critical path in planning processes like Airplane Turn-round in Chapter 6, the
‘goodness’ of washing machines in Chapter 7, and many more.
Other familiar measures include the energy ratings for electrical appli-
ances, various labelling systems for the healthiness of food products, and the
Body Mass Index, used as a measure of obesity, defined as
mass in kg
BMI =
(height in metres)2
Note that the dimensions of BMI are mass/length2 or density × length so, as-
suming the density and shape of human bodies is roughly constant, the key
DOI: 10.4324/9781003303503-13
Designing measures 227
variable in BMI has the dimensions of a length. Waist measurement, another

widely recommended measure, reflects this directly. The strengths and limita-
tions of these measures are widely discussed.
How do we ensure that a measure captures the essential features we want?
For some things it seems straightforward. Money is an obvious example:
the amount of money in your bank account is important and well-defined.
Though, even with money, our primary concern may be the ‘buying power’
of that money which changes with time. As discussed in Chapter 9, the
well-defined rate of interest on a savings account is less important from a
longer term perspective than the real rate of return after inflation is taken
into account. That requires devising a measure for inflation as discussed in
Section 12.3. For other abstract and social constructs, the amount of inequal-
ity in a country, for example, as in Chapter 8, it is far from obvious how a
measure should be defined.
Creating definitions is an important aspect of doing pure mathematics, but
often they are just stated and then learned definitions. Interesting fundamental
choices in designing a definition, that give insight into the reasons for rules, can
be all too easily passed over. Why, for example, is the product of two negative
numbers defined as positive? The compelling reason is to make the distribu-
tive law work for negative numbers, so that the usual arithmetic of positive
numbers continues to work in the extended system. For example, both of these
calculations give the answer 1 when ‘a minus times minus equals a plus’.
( 4 − 3)(3 − 2) = 1 × 1 = 1
and
( 4 − 3)(3 − 2) = 4 (3 − 2) + −3(3 − 2) = 12 − 8 − 9 + 6 = 1
In this chapter, however, we look at the design of measures of abstract or
social quantities of real-world interest that go beyond physical measurement.
Constructing such measures is in fact a challenge of mathematical modelling:
identifying the critical variables and choosing a function that combines them
to create a useful number. First in Section 12.1, we focus on some classroom
activities involving the design of measures, going on in Section 12.2 to a more
general discussion of desirable properties of a measure. Finally, in Section
12.3 we look in more detail at the mathematics of some of the important
measures we have discussed in earlier chapters.
12.1 Designing measures in the classroom

We begin with some tasks inspired by the Balanced Assessment Project de-
sign group at the Harvard Graduate School of Education, which called them
‘-ness tasks’. See Guadagnoli (2000) for examples of the Harvard Balanced
Assessment team’s work.
228 Designing measures
1 Order these rectangles from ‘least square’ to ‘most square’.

2 Now design a formula for a measure of ‘squareness’.
3 Compare your measure with other students’ formulas. Which do you think is
the better measure and explain why.
4 Try to devise a better measure.
FIGURE 12.1.1 ‘Squareness’ task.

Source: Inspired by a Harvard GSE Balanced Assessment task [12B].
In Figure 12.1.1, students are asked to design a formula for a measure of

‘squareness’ S and compare it with their fellow students’ suggestions. A pos-
sible discussion might be as follows:
Jo suggested the length (longer side) l minus the width w. S = l – w.

Samira suggested the ratio S = l/w.
Samira’s measure only depends on shape; Jo’s changes with size.
Students are then asked to revisit and improve their work. For example:
Samira’s measure, S = l/w. goes from 1 to infinity. Perhaps it is better seen

as a measure of ‘unsquareness’? An improvement might be S = w/l which
goes from 0 (super-thin, a line) to 1 (exactly square). More complex pos-
sibilities could be devised, especially when extending to shapes that are not
all rectangles.
The original Balanced Assessment materials include several tasks in this genre:
disc-ness is similar to square-ness but looks at three-dimensional cylinders and
compact-ness compares clusters of dots These tasks are abstract and ‘pure’
but help introduce some of the concepts about what makes a good measure.
The four maps of imaginary towns above are all drawn to the same scale (about
5 km wide) and have about the same population.
Each dot on the map shows where one person or family lives.
1 Look at the maps and put them in order of how “crowded” you think they look.
2 Now design a method for calculating a measure of “crowded-ness”.
3 Compare your measure with other students’ work. Which do you think is the
better measure and, most important, explain why.
4 Try to devise a better measure.
5 Extended task – find the population of some real medium-size towns and
cities. How many people might one of the dots on the map above really
represent? Look at satellite photos online and try to guess where most people
live – could you estimate a value for your crowded-ness measure?
FIGURE 12.1.2 A more practical -ness task [12B].
The task in Figure 12.1.2 (and [12B]) is inspired by Harvard’s Compact-ness

idea but with a slightly more realistic context. A possible solution might be
to calculate the average population density by dividing the number of dots
by the estimated area of the built-up zone … or should that be the area of the
map (which would be the same for all four towns) or the ‘bounding box’ of
the dots? Or is a town with large clumps of dots ‘more crowded’? Perhaps
you could divide the map up into a grid, count the dots in each square, and
take the mean/median/maximum…? The aim of this task is not to produce a
‘right answer’ but to provoke discussion of what a good measure is and how
widely used measures like population density should be interpreted.
Creating a measure for correlation

This lesson from the Mathematics Assessment Project [12C] is intended to
provoke students to think about how to measure the extent to which two
variables are related. The mathematics of correlation and least squares is
the professional answer to this question: these students approach the prob-
lem without that. The task uses the semi-plausible ‘real-world’ context of a
movie theatre owner conducting surveys for business planning – ice cream
sales versus temperature, snack sales versus soda sales, number of viewers
versus movie length. By happy chance, a quick glance at the scatter graphs
suggests that the ‘data’ for one is strongly correlated, one weakly correlated,
and the other is fairly random. Students are first asked to critique a sugges-
tion of drawing a polygon around the scatter points, calculating its area
and using 1/area to measure ‘correlation’. They are then asked to devise
a better method. In the follow-up lesson, they are given three examples of
‘student work’ (carefully chosen and tweaked by the designers) to compare.
Figure 12.1.3 shows extracts from two of these examples – the second could
be used to lead into teaching ‘least squares’.
‘student’ work used in the MAP Design a measure lesson.

FIGURE 12.1.3 Sample
Mathematics Assessment Project [12C].
12.2 What makes a good measure?

Even within the examples above, where the situation is simple and clearly de-
fined, there are design issues worth discussing. For more complex life-related
situations, as we have seen in previous chapters, these issues can be pro-
found. Here we stand back from particular contexts and take a more general
look at the properties that characterise a good measure, illustrating these
with examples from previous chapters.
A measure is stronger if it
• quantifies an important feature of the situation of interest,

• is well-defined and explained so that anyone can use it,
• takes values that reflect the extent of the feature in a way that seems natu-
ral, and
• uses data that is easy to collect.
BMI, for example, fulfils these criteria; that does not eliminate its
limitations.
There are also some other, less obvious, properties that are important.
For example, no widely adopted measure has just the use, and the influence,
for which it was designed – at the least, there are usually unintended conse-
quences. Indeed, one rather general problem is summarised in Goodhart’s
(1975) Law, usually simplified as follows:
When a measure becomes a target, it ceases to be a good measure.
This may be because people ‘game it’ – manipulate their activity pattern
to help meet the target, which may be good but may involve suppressing
other important activities not included in the measure. This is a design
issue – some important aspects of the real phenomenon being measured have
been overlooked or undervalued by the measure. Since the whole point of a
measure is to reduce a complex phenomenon to an easily compared number,
this is almost inevitable to some extent. Goodhart was originally critiquing
monetary policy, but the principle extends to many other fields. If you need
a minimum number of words in your essay, waffle. If you’re a call centre
operator ‘scored’ on how many calls you answer per day, get callers off the
line as quickly as possible without solving their problem. If you want to in-
crease your rating on social media, post lots of messages that a lot of people
will ‘like’ and so ‘upvote’. Criticising Android in an iPhone forum – or vice
versa – is usually a banker. There are many examples of this in education,
notably the effect of ‘high-stakes’ examinations on teaching, which we look
at in Section 12.3.
To counter this, a good measure needs an explicit ‘improvement mecha-

nism’ that evaluates the mechanism in use against the aims for which it was
introduced. Price indexes, for example, are based on a ‘basket of goods’ that
are representative of buying habits, so the goods included need to change
with time. As we noted in Chapter 8, this makes comparisons over long peri-
ods increasingly approximate.
Log versus linear presentations

There are many issues about the presentation of data, which are affected by
the design of the measure. Figure 4.4.4, for example, using data from early
in the pandemic, presented the risk of dying from COVID-19 as a function of
the patient’s age (the key variable) in both linear and logarithmic form. The
latter proved more informative because the risk changed by many orders of
magnitude, so the logarithmic scale could show all parts of the graph clearly.
As it happened in this case, there was an extra benefit from the logarithmi-
cally defined measure because, from middle-age upwards, the dependence on
age was close to exponential, the risk doubling every eight years of age, so the
graph is roughly a straight line.
Sometimes, logarithmic scales are a better model of human perception
of phenomena. For example, the musical scale of octaves is a logarithmic
one – each octave represents a doubling of frequency. Imagine listening to
someone playing each note on a piano keyboard from the low bass to the
high treble and sketching a graph of the ‘pitch’ you heard. You would prob-
ably perceive the pitch as rising steadily (so a straight line graph), rather than
the frequency’s exponential curve doubling every 12 notes. The decibel scale
of loudness is also logarithmic, as is the Richter scale for the magnitude of
earthquakes – both phenomena that vary over many orders of magnitude in
the energy involved.
An unusual logarithmic measure features in Chapter 13 which considers
how to compare the difficulty of guessing passwords creating using various
rules – say, ten random alphabetic characters compared to a random English
word or compared to a longer (but memorable) string of random English
words. Password entropy uses a logarithmic scale to estimate the number of
random binary bits (hard to crack, but impossible to remember) that would
give the same number of permutations.
12.3 Some important measures – a critical review

This section is largely focused on mathematics. The societal issues are dis-
cussed more fully in earlier chapters. We start with an example from educa-
tion, moving on to look at some important measures that we have discussed
in earlier chapters, some including more advanced mathematics.
Unintended consequences of ‘high-stakes’ assessments

In many countries governments introduce compulsory assessments of stu-
dent performances – examinations or tests – believing that they lead both to
improved learning and to decision-makers making better informed decisions
whether it be parents choosing a school or the government making policy
decisions. Surely an admirable goal. “If you can’t measure it, you can’t man-
age it” again. However, they also want assessments that satisfy constraints:
• standardised – in the interest of fairness, surely essential,

• well-defined assessment objectives,
• don’t take much “time away from teaching”,
• don’t cost much, and
• provide results that are clear and easy to understand (just one number
preferred).
The normal outcome is assessments that, particularly in mathematics, are

time-limited taking only an hour or two at most, consist of short ‘items’ that
take only a few minutes to answer, with no extended chains of reasoning or
student explanation. It is no surprise that teachers, whose career progress
often depends on ‘test scores’, focus on task types with the limited range in
the tests and often school leadership reinforces this. Burkhardt et al. (1990)
summarised this as:
What you test is what you get (WYTIWYG).
Sadly, these assessments do not encourage many of the dimensions of learn-

ing that are recognised as important, particularly those involving substantial
chains of reasoning or enquiry. While WYTIWYG has long been a familiar
part of teachers lives, government and assessment providers have been reluc-
tant to accept it as a fact. You hear phrases such as “We don’t test this but, of
course, all good teachers teach it” and “We’re just sampling some important
aspects of learning”. This allows them to avoid the responsibility to provide
‘tests worth teaching to’ that sample all the ambitious goals for learning that
are commonly listed in official curriculum documents. More recently, that
has changed in some places with assessment objectives that, for example, sep-
arate skills, reasoning, and solving non-routine problems. But such progress
in the rhetoric is, up to now, minimally reflected in the examination tasks
themselves. Traditional tasks that demand extended chains of reasoning with
explanations (commonplace as essays in humanities subjects) are now rare in
mathematics assessment in some anglophone countries. Yet, as we have seen
in this book, mathematical literacy largely involves such activities. PISA tasks
are an invaluable but limited step in this direction.
Assessment does not have to be narrowed in this way. Past examples show
(see, for example, Burkhardt, 2009) that even timed written examinations
can cover a much broader variety of mathematical reasoning. Extended pro-
ject work can be reliably assessed, although it takes time. Practical principles
for the design of High-stakes Examinations to Support Policy (Black et al.,
2012) were developed by a working group of the International Society for
Design and Development in Education.
Picking your inflation measure: RPI versus CPI

There is a nice mathematical issue behind the measure of inflation currently
preferred by the UK government – ‘Consumer Prices Index’ (CPI) – which,
since 2003, has been gradually replacing the previous ‘Retail Prices Index’
(RPI) measure. CPI is the international measure, used by the OECD and
governments to study inflation around the world. Both measures are cor-
rectly described as the percentage increase over the past year of “the average
price of a basket of products and services” that is chosen to represent the
spending of a typical citizen. (For the CPI, the actual goods and services are
country-specific.)
One interesting mathematical difference between the two indices is that
the ‘retail price index’ for inflation (RPI) uses the arithmetic mean of the in-
creases across the basket while the ‘consumer price index’ uses the geometric
mean of the increases. It is standard mathematics that the former is always
larger – for two numbers, a and b, the difference of the squares of arithmetic
and geometric means is the square of half the difference, which is always
positive (or zero):
 a + b2 
− ab = ( a + b ) − 4 ab = a 2 + 2ab + b2 − 4 ab = ( a − b )
2 2
4  
  2  
So the difference between the results is influenced by the difference between

the arithmetic and geometric mean of a sample, which in turn is proportional
to the variance of the sample – illustrated by ( a − b ) in the simple formula
2
for two items above.

The UK government still uses both CPI and RPI. Interestingly, it uses
• the lower CPI for increases in payments such as pensions, unemployment

and disability benefits;
• the higher RPI for increasing some charges, for example, train fares.
You may wonder why!

There are other important differences between RPI and CPI – notably
the baskets do not contain exactly the same types of products and services.
RPI includes housing costs (rents, mortgage, payments) which CPI does not;
CPIH, a version of CPI that allows for housing costs (by incorporating yet
another measure – ‘rental equivalence’) is currently being introduced in the
UK. A fuller discussion of RPI versus CPI can be found on the UK Office for
Budget Responsibility’s website [12D]. Despite their complexities, estimates
of inflation are nonetheless essential to the modern world.
Choosing the ‘basket of goods’ becomes even more complex over time.
Year-by-year an item ‘drops out’ as it becomes less significant in people’s buy-
ing habits; another item is introduced. Such gradual change is non-controver-
sial. But measuring inflation over decades, or even centuries as in Chapter 8,
where the pattern of buying has changed qualitatively is much more difficult.
The price of bread, for example, is much less significant in modern lives than
it was long ago, and travel much more important.
How can you measure economic inequality?

This is not an easy question and has no unique answer. For instance, do
you include wealth or only income? The measure used by Wilkinson and
Pickett in The Spirit Level which we summarised in Chapter 8 is the ratio
of the average of the highest 20% of incomes to the average of the lowest
20%; this varies in the data shown from below 4 in Japan to nearly 10 in
the United States.
A more sophisticated measure of inequality is the Gini Coefficient [12E]
– a number between 0 and 1 (though often expressed as a percentage) that is
roughly the average unsigned difference in income between pairs of individu-
als in the population divided by the average income, that is, for n people:
∑in=1 ∑ nj =1 xi − x j
G=
2n 2 x
The factor 2 arises because each pair is counted twice – each individual is
both i and j in the n2 terms. It follows that
• G = 0 if and only if everyone has the same income.

• G ≈ 1 if one person in a large population has all the income.
• G = 0.5 if half of the population have the same high income, and the other
half all have the same low income.
The Gini coefficient is not just for inequality of income. Clearly, wealth
could be used instead of income, but is also useful in other domains. For
example, Gini coefficients could be used to compare inequality in mathematics
achievement between individuals within or between schools. Results from
recent PISA studies (often using Gini coefficients) show that achievement
FIGURE 12.3.1 Visual representation of the Gini coefficient and Lorenz curve in

(a) general case and (b) special two income case.
Source: Adapted from illustrations on Wikipedia [12E].
is negatively influenced by academic and social segregation (inequality) be-

tween schools (OECD 2019), adding to the negative influence of individual
inequality.
Another representation is through the Lorenz curve, which shows the
cumulative income or wealth through the population, rank ordered from
poorest to richest – illustrated in Figure 12.3.1. The slope of the curve rep-
resents the income at that point in the population. If everyone has the same
income, the total grows linearly – the 45 degree line in Figure 12.3.1a. If
one person has all the income, the curve runs along the horizontal axis and
then shoots up the vertical axis. The curve shown represents a more typical
population.
Figure 12.3.1b shows a situation where there are just two different incomes
and the top u of the population has f of the income. The Lorenz curve consists
of two lines – one for the (1 − u) low paid folk, the other, steeper, line for the
high paid. We can calculate the ratio of areas in this case giving
1 1 1 
G = 2  1 − (1 − u) (1 − f ) − u (1 − f ) − uf  = f − u
2 2 2 
A surprisingly simple result. (When one meets a complex calculation with a

simple result like this, one should ask: “Can I find a simpler route to it?”)
Good to check that it fits the extremes cases: for equal incomes f = u = 100%;
G = 0; for one person with all the income f = 1, u = 0; G = 1.
12.4 The importance of units

Before we finish this chapter on measures, we should say something about
the units we choose to express quantities in. Visualising and communicating
the meaning of numbers effectively depends on understanding the units used.
It can be made much clearer by using standardised units.
History tells a relevant story. Before the French Revolution of 1789, there
were thousands of different units for length and weight in use across France.
When most craftsman never went far from their village, this did not matter.
The next century led to a government-led movement to standardise units in
a ‘metric system’. Initially physical objects – a standard metre bar and kilo-
gram sphere kept in Paris – were used to define the unit. Subsidiary standards
all over the country were then standardised against these – a cumbersome
method. At a convention in Paris in 1875, other governments were persuaded
to adopt the metric system.
Originally, physical ‘standard’ objects used had to be carefully main-
tained at national weights and measures bureaus. For example, until 2019,
the kilogram was still defined as the weight of a particular lump of metal –
the ‘International Prototype of the Kilogram’ held at the International
Bureau of Weights and Measures in France. Countries had their own copies
of this kilogram. The international standards movement has been work-
ing to redefine the standards (though not the size of the units) so that they
can be reproduced anywhere. The second is based on the frequency of the
light of a specific atomic spectral line and the metre is derived from that.
Since 2019, the seven ‘base units’ of the SI system have all been redefined in
terms of measurable physical phenomena. The best-known units are given
in Table 12.4.1, more detailed information can be found from the National
Physical Laboratory website [12F].
TABLE 12.4.1 Three of the base SI units
Unit Symbol Description Definition
second s Unit of time Defined as the time that elapses during

9 192 631 770 oscillations of light of
a specific spectral line of the caesium
133 atom
metre m Unit of length Defined by the value of the speed of light,
299 792 458 ms-1, and the definition of
the second
kilogram kg Unit of mass Defined by the value of Planck’s constant,
6.626 070 15 × 10−34 kg m2s−1, and the
definitions of ‘metre’ and ‘second’
Source: Based on information from the National Physical Laboratory [12F].
The value of using standard units for making comparisons is now gener-
ally recognised. For example, the way that national income or GDP should
be calculated is internationally agreed (though not universally observed by
governments!) In other areas there are avoidable distractions – for exam-
ple, during the COVID pandemic, death rates ‘per hundred thousand people’
were commonly used rather than per thousand or per million as international
conventions encourage.
As we have noted (see, for example, Section 3.3), understanding very large
and very small numbers is an important part of mathematical literacy – in
particular the extending list of names for the various powers of a thousand:
kilo, Mega, Giga, Tera, Peta going up and milli, micro, pico, femto going
down. Especially driven by computing technology, the range of numbers in
common use continues to increase.
Perhaps the most common cause of misunderstanding is confusing quan-
tities and rates – as with energy and power, capital and income, distance
and speed (familiar to mathematics teachers, though people are usually
clearer on this one). Discussions of power generation and use – in the con-
text of moves to mitigate global warming, for example – need to distinguish
the rate of energy production in power stations, wind turbines, or solar
panels (Gigawatts) from the household energy consumption that appears
on a monthly bill (kilowatt-hours), or its national annual equivalent. The
official SI unit of energy – the Joule – is largely neglected in this context (for
what it’s worth, 1 kilowatt hour is 3.6 Megajoules).
It is curious, and unfortunate, that experts are often sloppy about units –
to the point of being misleading. Discussions about GDP and national debt
are an egregious example. “This means the national debt will grow to 75%
of GDP” is a typical statement in the media. But the national debt is a lump
sum of money (or rather lack of it), whereas GDP is annual rate of expendi-
ture of money, so “9 months GDP” is what they mean. When challenged,
they say “Everyone knows what I mean”. They don’t. Nobody would say
“The distance from Liverpool to Manchester is 50% of the motorway speed
limit”.
Back to our general point: units matter. They clarify meaning.
Acknowledgements
Figures 12.1.1 and 12.1.2 were created by the authors but were inspired by
the work of the BA/MARS team at Harvard Graduate School of Education.
Figure 12.1.3 is by the Math Assessment Project team and appears courtesy
of Bell Burkhard Daro Shell Centre Trust. Figure 12.3.1 is adapted from a
public domain image from Wikipedia (left) and an image by ‘Woodstone’
(right) licensed under the Creative Commons Attribution-Share Alike 4.0 In-
ternational licence (https://creativecommons.org/licenses/by-sa/4.0/).
References
Black, P., Burkhardt, H., Daro, P., Jones, I., Lappan, G., Pead, D., & Stephens, M.
(2012). High-stakes examinations to support policy. Educational Designer, 2(5).
Retrieved 16 October 2023. Retrieved from http://www.educationaldesigner.org/
ed/volume2/issue5/article16
Burkhardt, H. (2009). On strategic design. Educational Designer, 1(3). Retrieved
16 October 2023. Retrieved from http://www.educationaldesigner.org/ed/volume1/
issue3/article9
Burkhardt, H., Fraser, R. E., & Ridgway, J. (1990). The dynamics of curriculum
change. In I. Wirszup & R. Streit (Eds.), Developments in school mathematics
around the world (Vol. 2, pp. 3–30). National Council of Teachers of Mathematics.
Drucker, P. F. (1995). People and performance: The best of Peter Drucker on manage-
ment. Routledge.
Goodhart, C. A. E. (1975). Papers in Monetary Economics, Volume I, Reserve Bank
of Australia.
Guadagnoli, T. (Ed.) (2000). Advanced High School Assessment, Package 2 (Balanced
Assessment for the Mathematics Curriculum). Dale Seymour Publications.
OECD. (2019). PISA 2018 Results (Volume II): Where All Students Can Succeed.
OECD Publishing, Paris. https://doi.org/10.1787/b5fd1b8f-en

mathlit.org – append the link code to go directly to the entry: –
purposes:
[12A] Measure (mathematics) – Wikipedia

https://en.wikipedia.org/wiki/Measure_(mathematics)
[12B] Example ‘-ness’ Tasks – mathlit.org
[12C] Designing a Measure: Correlation – Mathematics Assessment Project
[12D] The Long-run Differences between the CPI and RPI – Office for
Budget Responsibility
https://obr.uk/box/the-long-run-differences-between-the-cpi-and-rpi/
[12E] Gini Coefficient – Wikipedia
https://en.wikipedia.org/wiki/Gini_coefficient
[12F] The SI Base Units – National Physical Laboratory
https://www.npl.co.uk/si-units
13
MATHEMATICS FOR INFORMATION
TECHNOLOGY
Chapter 10 discussed some of the many exciting ways in which Informa-

tion Technology (IT) could support classroom teaching in ways particularly
relevant to mathematical literacy. Sections 10.5 and 10.6 started to look at
some examples of how the widespread use of technology was making new
demands on mathematical literacy. In this chapter we approach the more
complex question of what mathematics should students know about to help
them engage in more depth with the technology that is now part of everyday
life. Developments in computers, communications, and information process-
ing have been profoundly changing our private lives and workplaces since the
mid-twentieth century, and ‘computer science’ is now an important field of
both pure and applied mathematics, yet this has had little impact on school
mathematics curriculum. Here we aim to pick out a few examples where ac-
cessible and widely taught mathematics topics can be made relevant to the
challenges of an increasingly digital world.
There is an entire field of ‘discrete mathematics’ that emerged in the mid-
twentieth century and has grown in importance alongside, often in connection
with, computer technology. Broadly, this is summarised as ‘the mathematics
of countable sets’, in contrast to the more traditional ‘continuous mathemat-
ics’ of real numbers, algebra, infinitesimal calculus, and Euclidian geometry.
In more practical terms, discrete mathematics forms a broad umbrella for
topics such as the integers, formal logic, graph theory (‘systems of nodes and
edges’ – not illustrative charts), and discrete calculus (the study of incremen-
tal changes as opposed to the infinitesimal changes of traditional calculus) as
well as the fundamental theories underpinning computing and the study of
algorithms. A modest example of the difference between discrete and con-
tinuous approaches is using a spreadsheet to model population growth via an
DOI: 10.4324/9781003303503-14
Mathematics for information technology 241
iterative formula (Figure 9.3.1) instead of formulating and solving a differen-

tial equation. Introducing ‘discrete mathematics’ to the curriculum was part
of the ‘new mathematics’ courses of the 1960s, but it has not significantly
displaced traditional mathematics. Yet the computers we know were partly
conceived by mathematicians (most famously Alan Turing) and the close re-
lationship between computer programming and mathematics are dramatic
examples of how mathematics can ‘change the world’.
In this chapter, we start by addressing how computers solve various types
of problems and pick out a few examples of the mathematics behind comput-
ing, progressing from simple calculations using well-known formulae to the
topical field of machine learning. This is followed by a look at just two ex-
amples from the vital field of computer security – choosing a strong password
and a brief introduction to encryption, both of which are highly relevant to
users of current technology.
We don’t expect students to learn how to use and apply this mathemat-
ics in any detail. Instead, these sections highlight ideas that are valuable for
a mathematically literate person to ‘know about’ in a technological world.
Both ‘know how’ and ‘know about’ feature in the context-focused math-
ematical framework (see Figure 1.1.1). These ideas might form the basis of
mathematical discussions or even essay projects. They are not central to most
mathematics curricula, but some students will encounter them in various
digital technology subjects, so there are opportunities for teachers to make
interdisciplinary links. As we discussed in Chapter 1, this absence of ‘know
about’ seems to be a particular feature of mathematics education – geography
students, for example, aren’t expected to reproduce the chemical equations
for aluminium smelting in order to discuss why it is an important industry
in regions with cheap renewable energy – they can ‘know about’ the energy
demands without ‘knowing how’ they are evident from the equations. In
contrast, mathematical tasks at school almost always ask ‘know how’ ques-
tions involving using a learned procedure, solving an equation, or proving a
theorem. So this chapter aims to help readers ‘know about’ the mathematics
underpinning important aspects of IT in action.
We have deliberately not addressed the teaching of programming here.
Learning coding can be a valuable addition to the school curriculum – both
as a practical vocational skill and as a source of deeper understanding of both
technology and mathematics – but it is beyond the scope of this book. Real-
istically, acquiring any level of fluency in a particular programming language
would take too much time to sneak in to most mathematics courses. We have
included a couple of snippets of simplified code for the benefit of those who
might have some experience of programming, but where possible we have
turned to spreadsheets to illustrate points. We propose (see Section 3.3) that
using a spreadsheet well as an important mathematical literacy skill for most
people to have, to be developed partly within school mathematics.
242 Mathematics for information technology
13.1 Numerical calculations and accuracy

In the following sections, we look at some examples of how computers solve
– or, sometimes, fail to solve – various problems. Some tasks seem straight-
forward to tackle using a computer, because they have been analysed math-
ematically and reduced to a formula. For example, the quadratic equation:
ax2 + bx + c = 0
can be rearranged algebraically to give the familiar formula:
−b ± b2 − 4 ac
x=
2a
This can easily be coded using a high-level programming language, a spread-
sheet, or even some pocket calculators as something like Figure 13.1.1.
Most popular programming languages only differ on fine detail such
as what the positive square root function is called (usually something like
“sqrt”). Of course, the above is not a complete program – it doesn’t input
the values for a, b, c or output the results. Also, there are a couple of obvi-
ous special cases that the computer won’t handle unless it is explicitly pro-
grammed to spot them: if the equation doesn’t have any solutions (no real
roots), then b2 − 4 ac will be negative, and trying to take the square root using
FIGURE 13.1.1 Simple quadratic equation solution as code (top) and spreadsheet

(bottom).
Source: Spreadsheet and code available on the website [13A].
Decimal Binary
100s 10s 1s 8s 4s 2s 1s
0 1 2 1 1 0 0
10 + 2 = 12 8 + 4 + 0 + 0 = 12
FIGURE 13.1.2 The number twelve as decimal and binary.
the standard sqrt function will make the computer stop with an error, as will
dividing by 2a if a = 0, even though bx + c = 0 has a solution.
In writing such a program, we take for granted that computers can ‘just
do’ that sort of calculation. We won’t go into the details of how, behind the
scenes, everything ultimately gets reduced to simple logical operations on
binary 0s (‘off’) and 1s (‘on’). There are, however, some consequences that
can derail your computer program and cause it to crash or give inaccurate
results.
Computers represent integers in binary (base two). The number twelve (1210)
in base ten (decimal) becomes 11002 in base two (binary) (Figure 13.1.2).
What if we want to represent a fraction or mixed number in binary – say
twelve and a quarter? The principle is the same as decimal – but instead
of the place value of each digit after being one tenth of place value to the
left, after the ‘binary point’, each column halves its place value, so 12.2510 =
1100.012 (Figure 13.1.3).
Of course, as with the decimal system, some real numbers (e.g. fractions,
square roots) will require an infinite number of digits to represent accurately
in binary. Large base ten integers will need very large numbers of binary dig-
its. Digital computers have space for a limited number of digits (called ‘bits’
in binary). Hence, real numbers are generally represented as ‘floating point’
numbers consisting of an integer part multiplied by an integer power of 2.
This is the equivalent of scientific notation in decimal, where
(
0.000001234567 = 1.234567 × 10−6 in proper standard form )
−12
= 1234567 × 10
Decimal 12.25 Binary 1100.01

100s 10s 1s 1
10 s 1
100 s 8s 4s 2s 1s 1
2 s 1
4 s
0 1 2 . 2 5 1 1 0 0 . 0 1
12 + 2
10 + 5
100 = 12 1
4 8+ 4+ + 0
2
1
4 = 12 1
4
FIGURE 13.1.3 The number twelve and a quarter as decimal and binary.
So the 1234567 part (the mantissa or significand) could be stored as a seven-

digit integer and the exponent (−19) as a two-digit integer and it could all be
stored in ten digits. In binary it needs more space (bits).
0.00123456710 ≈ 0.0000000001010000111010001001100102
= 1.010000111010001001100102 × 210 − ten
= 1.010000111010001001100102 × 102 −1010 (binary)
However, note the “≈” in the first line – that is only an approximation
because most fractions that can be written precisely as decimal fractions are
recurring binary fractions. For example:
0.110 = 0.0001100110011001100110011001100110011…2
ACTIVITY
As in base ten, all fractions either terminate or repeat in binary. Investigate

which fractions terminate in binary and which repeat.
So, if you enter 0.110 (precisely 1/10), then behind the scenes it is getting stored
as a recurring binary fraction truncated to (usually) 23 or 52 ‘binary places’
and it is not quite equal to 0.110. That shouldn’t be surprising – many real num-
bers can’t be expressed precisely as decimal fractions either – but in a computer
program or a spreadsheet you can be fooled by entering what looks like a pre-
cise value in decimal resulting in ‘rounding errors’ that become significant over
multiple operations. Figure 13.1.4 shows this happening in a spreadsheet –
after multiple additions of 0.01, the result is very slightly wrong. This looks
innocuous (you’ll have to increase the number of decimal places to the maximum
to see it) but the result is that the value never reaches exactly 2346. This sort
of thing can cause problems for programmers who assume exact results. The
simple program in Figure 13.1.5 starts with f = 2345, repeatedly adds 0.01 and
is told to stop when f = 2346. We intend it to reach 2346 after 100 iterations
and then stop. But this program will likely fail and keep running indefinitely
because, like the spreadsheet, the closest f gets to 2346 is 2346.00000000002
which is not equal to 2346. So, when programming with floating point num-
bers, it is very important to avoid checking that two numbers are exactly equal,
or to rely on the result of subtracting two closely equal numbers.
Let’s say you fix the program in Figure 13.1.4 to stop when f ≥ 2346, so
it does actually finish – it then prints 2.1827872842550278 × 10−11 instead
of 0. That’s a very small error but your programming language was meant to
be accurate to 15 significant digits. This demonstrates how rounding errors
accumulate in complex programs.
FIGURE 13.1.4 Rounding errors in spreadsheets (top) and code (bottom).

Source: Spreadsheet and code available on the website [13B].
The limited precision of floating-point numbers can trip up programs like

our quadratic equation calculator if we want high accuracy. Say we wanted
the computer to solve:
x2 999999999x 1
+ − =0
10 000 100 000 000 1000
 1 
 Note: roots x = and x = −100 000

10 000
A computer algebra system will give these two roots precisely, because it solves
the equation analytically, by rearranging and factorising, and will keep the frac-
tions as fractions where possible. However, our spreadsheet (see Figure 13.1.5)
FIGURE 13.1.5 Spreadsheet quadratic solution showing rounding error.

FIGURE 13.1.6 A more accurate spreadsheet quadratic solution.

gives the roots as 9.99999993922529 × 10−5 and −100 000 – again not the
15 significant digit precision we were expecting. To make it worse, we could
have guessed x ≈ 0 if we wanted an approximation. The problem is, in our
quadratic formula, if ac is much smaller than b, then b2 − 4 ac is going to
be very close to b and so calculating one of the roots will involve subtract-
ing two very similar numbers, and we have seen from Figure 13.1.4 that this
can leave a result which is dominated by rounding errors. The moral is that
you can’t always just translate a mathematical formula into your favourite
programming language or spreadsheet and expect precise results. Instead, it
is sometimes necessary to rewrite the formula to avoid rounding errors in
intermediate results – see Press et al. (1986, p. 145). A better – but harder
to follow – spreadsheet to solve our equation is shown in Figure 13.1.6. This
appears to just be an unnecessarily round-about way of calculating the same
thing, but which avoids subtracting b2 − 4 ac from b where the two are very
close and the difference would be affected by rounding errors. Note: Row 8 in
the spreadsheet contains a conditional function needed to check whether b is
positive or negative and choose one of two values for the intermediate value q.
Evaluating mathematical functions and solving general equations

We take for granted that computers know things like square roots, loga-
rithms, and trigonometric functions – but how? They do not always have
tables of values stored inside but use inbuilt rules that calculate each value
when required. These rules can only use basic operations such as addition and
subtraction, multiplication and division. The mathematics behind this is be-
yond any expectations of mathematical literacy, and the algorithms actually
used are heavily optimised for speed. However, some of the basic concepts
feature in more advanced high school and college mathematics courses, and,
when they do arise, it is worth making the connection to computing as an
important practical application.
Taylor’s series (and similar series expansions) can turn functions into an
infinite series of terms, for example:
x3 x 5 x 7
sin ( x ) = x − + − +
3! 5! 7!
The first terms of this series could be used to estimate the sin() function us-
ing just multiplication, addition, subtraction, and division – the more terms,
the greater the accuracy. Computers probably do not use this exact method,
for one thing it is probably too slow, but the idea helps understand how it
is possible for computers to ‘know’ these functions. An example spreadsheet
calculation using this expansion is shown in Figure 13.1.7. The value of x is
entered in cell E1. Column A gives the position numbers of each of the (first
seven) terms. Column B calculates the odd numbers for the denominators,
and Columns C and D enable calculation of each term. In column E, the val-
ues of the terms are successively added. The example shows a good estimate
for sin π /6 = 0.5 is given with just a few iterations.
Some other mathematics topics from college and higher level secondary
school classes are related to algorithms that are built into computers (in an
optimised form):
• Newton’s method to improve an estimate of the solution to an equation.

Computers often use this and similar methods to calculate square roots.
• Trapezoid rule for estimating integrals (likewise, estimating the first de-
rivative from two closely spaced points on a curve). The easiest way for
computers to do calculus.
FIGURE 13.1.7 Taylors series expansion of sin(x) using a spreadsheet.

Source: Spreadsheet available online [13D].
• Method of differences can make summing series far more efficient. This is
significant to the pre-history of computing, as Charles Babbage’s Differ-
ence Engine was a mechanical calculator intended to use this method to
calculate scientific tables. The Difference Engine was never completed, but
subsequent mechanical devices used the same theory.
13.2 Storing and organising data
Sorting
One of the most widespread uses of computers in business and commerce
is storing and organising large databases. A database needs to be easy to
update, easy to keep consistent, efficient to search, and capable of producing
‘reports’ which summarise or reorganise the data in useful ways.
A common task in data storage is sorting lists into numeric, alphabetic,
or other order – a much-studied subject in computer science on which whole
books have been written (famously Knuth, 1973, weighing in at 723 pages –
plus fold-out!). If you are sorting millions of items, having the most efficient
algorithm can be vital.
How would you instruct a computer to sort items? You could experiment
with ‘rules’ for sorting using, say, some letter tiles from a word game:
TKPEYOXSA
Looking at those, you would immediately start thinking “oh, X and Y go at the
end, the A on the front, then the E” and the job would be half done. But what
if you (like a computer) didn’t have that sort of human insight, and could only
pick two letters from the list and ask which one should come before the other?
One algorithm – called a ‘bubble sort’ – is particularly easy to understand and
illustrates the sort of logic that can be simply programmed into a computer.
1 Work through the list from left to right, one at a time and compare each
letter with the following letter.
2 If the two letters are in the wrong order, swap them.
3 Keep going until you’ve checked the last two letters.
4 If you had to make any swaps, go back to step 1.
5 Keep going until you make a pass through the list without making any
swaps – your list is now sorted! No cheating – with these rules the com-
puter wouldn’t notice that the job is finished until it had gone through the
list comparing every pair of letters that one last time.
It is worth noting that this is a horribly inefficient process, as you’ll see

if you try to follow it by hand. In the example you need 8 passes to get the
A from one end of the list to the other, meaning 64 pairs of letters to com-
pare. Even without such a ‘worst case’, this is what is known as an O(n2) or
“order n2” algorithm, because the number of steps tends to go up with the
square of the number of elements in the list. This is not a problem if you’re
sorting a handful of items, but it is a problem when there are millions of
items. You can probably work out a few ways to make the sort more efficient.
Today, most programming languages have more sophisticated sort routines
built-in and ready to use. These are too complex to describe to a human
but can sort a list using the order of n log n operations. Most programmers
wanting a simple sort would use an existing tool – the above examples are
intended as an illustration of how computers tackle this sort of straightfor-
ward but laborious tasks, and to raise awareness of the attention that goes
into doing apparently straightforward tasks efficiently.
Efficient searching, ‘keys’, and relational databases

One benefit of sorting a list of items, as discussed in the previous section, is
that it is then easy to search, say, a list of names using an efficient “binary
search” without having to check every single list item.
1 Look at the value in the middle of the list.

2 Does it come before or after your target name in alphabetic order?
3 If before, discard the top half of the list. If before, discard the bottom half
of the list.
4 Repeat until you find the target.
In a large database, however, the entries are often not in any particular
order, and there may be several different things you may want to search for.
You can’t resort whole database for every different search. The solution is to
produce one or more ‘indices’ or ‘keys’. Figure 13.2.1 gives a trivial example
FIGURE 13.2.1 Database of car licence plates with two indices.

Employee ID Name Date of Birth Room # Desk # Car licence plate

1 Harold 21/3/1987 3A 1 WPQ 9210
2 Chris 14/6/2002 3A 2 XGR 3451
3 Krishna 3/7/1999 3B 1 TPK 1231
4 Asha 9/8/2001 3B 2 WHT 4812
5 Nile 17/10/2000 4 1 KRD 5674
FIGURE 13.2.2 A flat file database.
of a database of car license plates with one index to help find entries by name
and another to search by license number. Each index is just a list of numbers
that point to entries in the main database, sorted in alphabetic order of
owners’ names or car license numbers, respectively, which can be used to
do a rapid binary search as described above.
A development of this idea – the relational database – is used in most large
business databases today. There is a sophisticated mathematical theory be-
hind relational databases, but one fundamental idea is that every important
piece of information should only be stored in one place, making it easy to
update and hard to add multiple, conflicting entries. Imagine we are trying to
extend the database in Figure 13.2.1 to keep track of employees, their per-
sonal info, their cars (for keeping track of parking), and where they sit. We
could extend it as in Figure 13.2.2. This is fine, until ‘life happens’. Chris has
a second car that they want to park sometimes (and, by the way, their name
is spelled “Kris”) and then Shana joins the company part time working
Monday–Wednesday sharing Nile’s desk (since he only comes in on Thursdays
and Fridays) and room 3A gets renamed “The Burkhardt Suite”. So pretty
soon you end up with something like Figure 13.2.3, with duplicated entries,
contradictions, and non-standard annotations.
Employee ID Name Date of Birth Room # Desk # Car licence plate

1 Harold 21/3/1987 Burkhardt 1 WPQ 9210
2 Chris 14/6/2002 3A 2 XGR 3451
3 Krishna 3/7/1999 3B 1 TPK 1231
4 Asha 9/8/2001 3B 2 WHT 4812
5 Nile 17/10/2000 4 (Thurs–Fri) 1 KRD 5674
2 Kris 14/6/2002 Burkhardt 2 TYB 9876
6 Shana 12/2/1998 4 (Mon–Wed) 1 none
FIGURE 13.2.3 Flat file database 6 months later.

Of course, if we were only keeping track of half a dozen employees, we

would be massively overthinking this. In reality, the database would be for
a firm with hundreds of employees and probably several people responsible
for updating the database, so the mistakes wouldn’t be so obvious, and (say)
trying to trace an improperly parked car or discover which desks were free
on particular days would become error prone. What is needed is a database
that makes it hard to enter invalid data, easy to update, and is easier to re-
structure when new requirements (such as desk sharing) arise. If we look at
our example, we can make some rules about the data.
• Every person has exactly one name, date of birth, and can be allocated a
unique employee ID which shouldn’t need to change.
• Not all employees have cars but some have more than one.
• Each car is owned by exactly one employee (we can worry about car shares
later – the car will still have one owner).
• Each car has a unique licence plate number and no two cars should have
the same number.
• Employees can share desks, but only one employee can use a desk on a
particular day.
• Each desk is in a room, and has a number that is only unique within that
room.
• Each room has a name.
Now we apply the principle that each bit of important data should be
stored only once. So, we could organise our (original) data as in Figure 13.2.4.
The data is now spread across several tables. “Employees” contains just the
FIGURE 13.2.4 Relational database.

personal details of each employee. “Cars” just contains car licence plates
and the ID of the employee they belong to. So, you could look up a car li-
cence plate, find the employee number and then look up the employee’s ID
in “Employees”.
That seems overcomplicated, until you add Kris’ second car. All you need
to do is add the license plate number and Kris’ employee ID to the ‘Cars’
table and you’re done. Looking up either license plate in “Cars” will give
you Kris’ ID, and looking up Kris’ ID will find both cars belonging to Kris.
Meanwhile, there’s still only one “Employees” entry for Kris, so we don’t end
up with two copies of their personal information which could get out of step.
If Kris gets rid of the second car, you can just delete it from “Cars” without
breaking anything. So we have a “one to many relationship” – one employee
can have many cars.
The shaded column in each table of Figure 13.2.4, which provides a
unique, unchanging ID for each entry in the table and which can be refer-
enced by other tables, is called the ‘Primary Key’. Database software can
ensure the integrity of keys by, for example, refusing to add a new row with
a duplicated primary key, or deleting a row with a key that is referenced by
another table. (In practice, a database designer would probably give each
table an extra column of guaranteed unique values as a primary key rather
than relying on externally supplied values like Employee IDs and car licenses
to be unique.)
The “Rooms” table gives each room a name. “Desks” assigns each desk
an ID and links it to a room and “Seating” assigns each employee to a desk.
So, when you rename Room 3A, it only takes one change to the row for
Room ID 1 to change the name for everyone.
That looks really overcomplicated until Shana wants to share a desk with
Nile. Then it means we could simply add a line to “Seating” to link Shana’s
employee ID to desk ID 5 (and accept that each desk could have more than
one user). If we wanted to keep track of who used what desk on what day,
you could add a day-of-week column to the “Seating” table. For each em-
ployee ID, there would be one entry for each day of the week that they used
a particular desk.
The specifications of these tables would be input to a database manage-
ment system. The details would include rules about which columns refer-
enced other tables, which columns (or groups of columns) always had to
contain values that were unique within the table, which could be left blank,
etc. The database management system would refuse to make changes that
broke the rules.
All this is far more complicated than just typing the information into a
spreadsheet. In fact, a database like this would usually require a program-
mer to design forms and ‘helper code’ that made it easy for the users of the
database to enter and update information without worrying how it was

split up between tables. However, a well-designed database like this will
help ensure that the data is always consistent and any updates are prop-
erly applied. Spreadsheets are fine until they grow over the years and be-
come unmanageable and full of errors. Relational databases, as described
here, are not the only type of database, but they have been a mainstay
of commercial computing since the late 1980s and most of the world’s
data management systems in government, healthcare, and business run
on them.
13.3 Heuristics and ‘calculated guesses’

You would usually assume that a computer program was designed to reli-
ably produce the “correct” result. True, a program based on a mathematical
formula will have limitations on the range of parameters it can accept and
the mathematical precision of the result, while a sorting program could be
overwhelmed with too many items, or fed with characters that it couldn’t
deal with, but fully developed versions of those algorithms would detect and
warn of such problems. In theory, within those constraints, and subject to
any assumptions and simplifications made in formulating the model, the re-
sults will be provably correct.
This is not always the case. Not all problems can be reduced to a math-
ematical formula or a simple, reliable set of steps. Other methods are
needed to tackle problems that are too difficult to fully analyse or where
computing the exact result would take too long or indeed may be impos-
sible. In fact, Turing and others proved mathematically that there are
some problems for which it is theoretically impossible to precisely com-
pute the solution. Inconveniently, the proof does not say which problems
are like that.
For problems which cannot be analysed in practice, it is often possible to
use a heuristic method to produce a “good enough” result by trial and error
or by applying rules of thumb. These rules do not fully describe the problem,
and there is no guarantee that the result will absolutely always be correct. But
it is all we have. Although the first example below is a game with no serious
real-world use, it illustrates a distinction which has important implications
for many real-world applications of computers which rely on such heuristic
methods and can never guarantee a correct answer.
Playing games
Take a game like noughts and crosses (also called tic-tac-toe)– how could you
program a computer to play it? Figure 13.3.1 shows the eight steps of a game
FIGURE 13.3.1 A typical game of noughts and crosses (tic-tac-toe). Both players
are following the rules described in the text.
that ends in a draw. If you’ve played the game more than a few times, you’ve
probably learned a strategy something like:
1 If your opponent can win on their next move by completing a row or di-
agonal, block them!
2 Otherwise, if the centre square is available, take that.
3 If not, and a corner square is available, take that.
4 Otherwise, pick any free square.
That would form a set of “heuristics” that could easily be turned into a
computer program. It would play a good enough game to beat someone with
no experience, but it is not guaranteed to win or draw every time. In fact,
noughts and crosses is sufficiently simple that it is a “solved game”. It is quite
practical for a modern computer (or even a good high school student) to ana-
lyse all the possible moves from a given point, pick the best outcome, and
play a ‘perfect’ game, guaranteeing to, at least, draw. However, the four rules
above – plus, maybe remembering a few ‘1 move to win’ positions – are prob-
ably closer to how a human would play the game.
ACTIVITY
How many possible games of noughts and crosses are there?
(An upper limit to the number of possible games is 9 possible positions for the
first move, 8 for the second and so on, giving 9 × 8 × 7 × 6… × 1 = 9! = 362880.
Deeper thought shows that is an overestimate as, for example, some games
don’t continue until the board is full. And appreciating symmetry shows that
there are only three really different first moves: centre, corner, or side.)
For a more complicated game, such as chess, there are many more pos-
sible games (about 5 × 1044 board positions estimated to give about 10120
possible games, versus ~105 for tic-tac-toe) so a complete solution (an algo-
rithm guaranteed to win or draw every game) is still beyond the power of
current computers. Instead, they combine looking a few moves ahead and
applying ‘heuristic’ rules or looking for correspondences with famous games.
No chess-playing program is mathematically guaranteed to win every game.
Yet, the top chess programs have now beaten grandmasters – and even years
before that had happened, chess-playing programs could usually thrash a
casual player. These are still successful commercial products.
Playing games may not seem like a ‘practical’ way of using computers, but
games provide insight into how computers solve other problems. Enumerat-
ing possible outcomes or factors and ‘scoring’ the results (in the noughts and
crosses, the scores are win, draw, and loss), while looking for shortcuts to re-
duce the number of paths that need to be explored is a widely used approach
in computer ‘intelligence’. These methods apply from medical assessments
to deciding whether you get a bank loan. The take-home message for math-
ematical literacy is the important distinction between a computer solving
a problem in the rigorous mathematical sense versus being able to tackle a
“real” problem as well as, or better than, a typical human by making what is
essentially a ‘calculated guess’.
Monte Carlo methods

Another method of finding approximate solutions problems that may be too
complex to fully analyse mathematically is to test it for a random sample
of values and use probability to estimate the result. These are called Monte
Carlo methods, named after the municipality in France that is famous for
having many casinos. For example, Figure 13.3.2 shows how a Monte Carlo
method can be used to estimate the area of an irregularly shaped oil spill by
picking 200 points at random on the image and calculating the proportion
that are inside the spill. This gives an estimate of the probability a randomly
chosen point will be inside the oil spill which in turn estimates the ratio of
area of spill to the area of map.
This example comes from the Australian Academy of Science’s reSolve:
Maths by Inquiry materials [13E] and is suitable for classroom use – in
the lesson, students compare the accuracy of the Monte Carlo method with
other geometrical ways of estimating the area. Other methods – such as
splitting the shape into triangles – struggle with an irregular shape such as
the slick shown. Even sampling a regular array of points risks missing some
features which fall between the grid. Randomly chosen points might also
miss important features, but if a lot of points are chosen, the chances of that
are small.
FIGURE 13.3.2 Finding the area of an oil spill by a Monte Carlo method.
Source: From reSolve: Maths by Inquiry & CODAP by The Concord Consortium [13E].
As with any classroom-friendly activity, the example is sufficiently simple

that a modern computer could solve the problem by “brute force” simply
counting each pixel in the image – but in more complex real-world applica-
tions, or where calculations have to be made rapidly (checking products roll-
ing off a production line, for example), such methods can greatly reduce the
number of data points that have to be tested.
13.4 Neural networks and machine learning

One way of solving a problem that cannot be completely analysed using
mathematics is to use a “neural network”. These are widely used in mod-
ern artificial intelligence and machine learning techniques. Neural networks
mimic the way that biological brains are believed to work (greatly simpli-
fied!). Rather than being built to solve a problem using mathematics and
logic, they are “trained” to produce the correct result using a set of known
examples.
Figure 13.4.1 shows a diagram of a simple neural network. Each of the
circles in the diagram represents a “neuron” with multiple input values and
a single output. The network starts on the left with the initial input values,
and moves right working through the various layers (columns) of neurons to
the output.
Each input to a neuron has an associated ‘weight’ (illustrated here by the
thickness of the lines). The neuron works by multiplying each input value
FIGURE 13.4.1 Simplified neural network.
by its weight, adding together the results and then applying an “activation
function” to calculate the output value (this function could be as simple as
“if total is greater than 0 then output ‘cat’, else output ‘dog’”). This output is
then passed on from that neuron to all of the neurons in the next ‘layer’ that
it is connected to. In Figure 13.4.1 all the neurons are connected to all in the
next layer, but this is not usually the case.
This network takes four numbers (or a four-digit number) as an input
and produces a single output. You wouldn’t normally use a network to do
arithmetic, but the inputs could be (say) a person’s height, weight, waist
measurement, and shoe size gleaned from their online shopping habits, and
the required output could be their approximate age, propensity for a certain
illness, or the probability of them being interested in buying a smartwatch. A
network of nine inputs could be the state of a noughts and crosses board, and
the output would be the next move. A larger network, maybe with hundreds
of inputs and more layers of neurons, might take the pixels of an image and
output a code giving which letter of the alphabet it ‘saw’. An image recogni-
tion system might reduce the image of, say, a face to a single, unique number
representing that face, regardless of the exact lighting, angle, and framing in
the image.
In any case, the network starts off by producing nonsensical output and
must be ‘trained’ on a set of inputs with known answers. This is done by
inputting an example, comparing the output to the correct result, and then
tweaking the weights of each neuron’s input accordingly – and repeating
this many times with a large set of examples until the results have the re-
quired degree of accuracy. The process can usually be done automatically
given a database of examples and a carefully designed training algorithm.
Once trained, not only should the network produce the correct answer for
any of the example inputs, but the goal is that it should work for inputs
that it hasn’t been trained on. This has to be verified by testing. Note
that words like ‘training’ or ‘learning’ are widely used for this process
but shouldn’t be taken too literally. The design of the training algorithm
is critical and its aim is to efficiently search for the optimum values of the
weights.
The rapidly growing field of machine learning often uses neural net-
works, as well as other techniques that depend on training a computer
on a bank of known data. The problem could be anything from targeting
advertisements, recognising faces, diagnosing diseases, or making a self-
driving car recognise hazards. It is important to understand that – even
when advertised as artificial intelligence – there is no conscious under-
standing involved and no guarantee that the responses are correct. While
there are ways of producing data to justify why a machine learning model
produced a particular – possibly incorrect – response, this is usually much
harder than for a more traditional model, where you might be able to
point to a line of code, or an assumption made by the modeller to explain
the result. Recently, my photo software automatically classified a photo
I took of a white fungus as a cat. It would probably be impossible to
describe to a human the features of the photo’s pixels which caused the
algorithm to select ‘cat’. So it is especially important that users of such
systems are alert for implausible results and also understand the math-
ematics of false positives and negatives and how they can sometimes lead
to counterintuitive situations. This is discussed in more detail, in the con-
text of medical testing (in which machine learning solutions are already
playing a role), in Section 4.5.
Systematic errors have already been found to arise in machine learn-
ing systems due to a bias in the data used for training. For example, a
machine learning system might be used to identify good applicants for
places at a university, with the hope that it will be able to avoid the biases
that human selectors implicitly apply. But if it is trained on the dataset
of previously successful applicants, the biases of previous selectors will
be built in. Such problems are exacerbated by the difficulty of justifying
why a machine learning system made a particular decision – especially in
lay terms – when there is no single rule that the developer can point to or
simply correct.
13.5 Picking a password

Today, many websites, phone applications, video and music services, and
other tech activities require you to create an account and choose a password.
You soon end up with dozens of passwords to remember. The easiest thing is
to use the same password for everything, but that is risky since if an attacker
manages to guess or steal your password for one service, they have access to
all of your other accounts. Even if you don’t choose an easy-to-guess pass-
word, it only takes one incompetently designed website, a ‘shoulder surfer’
or being fooled by a ‘phishing’ email for it to leak out. So really, you should
choose a unique, clever password for every service you use and remember
them all.
How do you choose a good password? There are several requirements.
1 The password has to be hard for a human to guess – so, no mother’s

maiden names or old chestnuts like “password”, “secret”, “12345678”
or “swordfish”. You can always find out the ten most commonly used
passwords with a web search.
2 It has to be hard for a computer to work out by brute force (just testing
everything), which is a lot to ask given the speed of modern computers.
Moreover, a hacker using a computer to break a password can be smart
by focusing their search first on common names or words, and variations
such as having ‘i’s and ‘o’s turned to ones and zeros.
3 You have to be able to remember it and type it quickly. Something like
“Qz$t17!pTT” would be a good, strong password that is highly unlikely
to be guessed, but not easy to type or remember. This means it would
likely end up written down.
One mathematical measure of password strength, taken from information

theory, is information entropy. For a password this is calculated as
log2(number of possible passwords)
with the unit “bits of entropy”. This measure is useful for comparing rules
for valid passwords, or methods of choosing memorable passwords. It is
those rules that determine the number of possible passwords to use in the for-
mula. Why use this measure? Say it was possible to use a completely random
64 digit (bit) binary integer as a password: the number of possible passwords
would be 264 – so it would take up to 264 systematic guesses to be sure of find-
ing the password. Pretty secure. The entropy of such a password would be
log2(264) = 64 bits. So the entropy for different password schemes tells us the
number of random bits that would provide equivalent security. Another way
of interpreting it is that that chance of guessing a password with 64 bits of
entropy on the first try is the same as getting 64 successive heads by tossing
a fair coin.
In reality, your choice of password will be more constrained – long binary
numbers would be tedious to type and hard to remember. A totally random
8 letter password made from the 26 letters of the alphabet would be one of
268 possible passwords and so have an entropy of log2(268) = 37.60 bits.
Not as good as our random 64-bit integer. But eight truly random letters
are still hard to remember. An eight-letter English word would be easier to
remember – but also easier to guess – since a computerised password guesser
will doubtless run through the dictionary before it starts trying every random
permutation. There are about 80,000 eight-letter English words, so the en-
tropy is only log2(80000) or about 16 bits, and of course, people only know
a fraction of those words, reducing the entropy even further. Of course, both
of those calculations assume that the guesser knows your password is eight
letters long and that it is an English word.
Note that we are making the reasonable assumption that a hacker will run
an organised search, trying dictionary words and well-known password
schemes (pa55w0rd, password42, #password etc.) before resorting to a
‘brute force’ search along the lines of AAAAAAAA, AAAAAAAB…
ZZZZZZZZ. Entropy is more of a rule of thumb guide rather than a precise
measure of ‘security level’.
ACTIVITY
What would be the entropy of a password selected by picking a lowercase

English word between 5 and 10 letters long? How does this compare with a
random lowercase password between 5 and 10 letters long? How would
allowing any mixture of upper- and lowercase letters change these entropies?
Number of letters Approximate number of

English words in the dictionary
5 158 000
6 20 000
8 80 000
9 41 000
10 35 000
Studying password choice and strength is an interesting mix of math-

ematics and psychology. An important finding is that the usual pass-
word rules can actually make passwords more predictable. So, requiring
a ten-character password to include at least one digit 0–9 in addition to
the 26 letters of the alphabet would appear to increase its entropy from
log2(2610) = 47 bits to approximately log2(3610) = 52 bits, assuming that the
password consists of ten purely random characters. A human being is likely
to start with one of a few thousand common five- to eight-letter words
(only 10–12 bits of entropy) and then use one of a handful of strategies
(change the ‘I’s to 1’s, the ‘o’s to zeroes, append a digit). So, instead of just
checking for the 2000 most common words, a hacker needs to check for,
maybe eight variations of each of those words. This sounds significant but
only adds 3 bits of entropy.
log 2 ( 2000 × 8) = log 2 ( 2000) + log 2 (8) = log2 ( 2000) + 3

Using two common words actually increases the entropy from log2(2000) to
log2(2000 × 2000) – that is, it doubles the entropy. A better strategy for a
memorable password might be to stick to the letters a–z and string together
several common words to make a nonsensical – but memorable – phrase.
This was famously (at least within IT culture) illustrated by Randall Munroe
in the XKCD comic strip [13F] who illustrated how a passphrase like
‘correcthorsebatterystaple’ was both more secure and easier to memorise
than a single word mangled to something like ‘Tr0ub4dor&3’ to meet typical
password rules.
ACTIVITY (FOLLOWING ON)
Study the cartoon in Figure 13.5.1. What assumptions have been made to esti-
mate the entropy of both “styles” of password?
FIGURE 13.5.1 Choosing a password.

Source: From xkcd.com [13F].
Another solution to the “strong but memorable password” conundrum

is to use a password manager application. The software equivalent of a key
safe, these applications will generate unique, hard-to-guess (but equally
hard to remember) random passwords for every website or service you
use, and store them securely, so you only have to memorise one master
password. That is the one to unlock the password manager. You can even
use fingerprint or face recognition to quickly unlock it without typing a
password. Of course, this means you have a file containing all of your
passwords sitting on your computer and, in fact, you’ll often want to keep
it online so you can use it from your mobile devices and appliances. This
would be disastrous in the wrong hands. There is a nice risk/benefit bal-
ancing issue there. Does the risk of having your passwords stored in a file
outweigh the benefit of having strong, unique passwords for all your on-
line activity? Of course, it would help greatly if the password file could be
so strongly encrypted that it was useless to anybody without your master
password.
13.6 Cryptography
Sending secret messages has always been of interest to people working with
highly sensitive information, but not so much in everyday life. (Some people
just find it fascinating.) The internet has changed that. The way it works is
very open and public, and it is not hard to intercept messages. Even if you’re
not plotting to overthrow the government, you might want to send your
credit card number to an online shop without the whole world being able to
intercept and read it. While you don’t need to be a cryptography expert to
use the sort of encryption software now built into web browsers and the like,
understanding some of the underlying ideas may help you understand the
benefits and risks of secure communication.
Many people will be familiar with a simple substitution cypher with vari-
ous different rules– change all the ‘A’s to ‘C’, all the ‘B’s to ‘Q’s, and so on, or
move all letters on by three places, wrapping round from Z to A. A popular
school mathematics activity is ‘cracking’ such cyphers by looking for clues
like letter frequency. This shows how weak such cyphers are. A much better
scheme is to use a longer keyword so that how each letter is encoded depends
on its position in the message. Figure 13.6.1 shows the basics of a keyword
cypher – real-world implementations would use a longer key phrase and add
extra layers of sophistication to how it is combined with the text, and can be
very hard to crack if you don’t know the key.
In Figure 13.6.1, the message is to meet in the hall at eight and the key
word is ‘wibble’. We assign numbers to letters, A=0, B=1, C=2… ignoring
spaces, punctuation, lowercase, and so on. We start with the text letter M,
letter 12, corresponding to the key letter W (letter 22). We add the key to the
FIGURE 13.6.1 Encoding with a secret key.

Source: Spreadsheet available online [13G].
text using modulo (clock) arithmetic base 26, giving 12 + 22 ≡ 34 ≡ 26 + 8

(mod 26). By adding the code for ‘M’ (12) to the code for ‘W’ (22) and work-
ing modulo 26 wraps around the alphabet to ‘I’ (8). To decode, we reverse
the process: start with I (8), subtract W (22), add 26 to avoid negatives, with
result 8 – 22 + 26 = 12, the letter M.
The problem with this system is that, somehow, both the sender and re-
cipient have to agree on the key and keep it secure. Imagine you want to send
your credit card details to an online shop without them getting intercepted.
You could encrypt them with a keyword, but then you have to send that to
the shop without it being intercepted. This is the problem with ‘symmetrical
encryption’ where the key to encode the message is the same as the key to
decode it.
The solution is ‘asymmetrical encryption’ which uses a pair of keys: one
to encode the message and one to decode, which allows what is known as
‘public key encryption’ (PKE). Using the credit card example, it works like
this.
1 The shop sends you the encryption key (their public key) – but keeps the
decryption key (their private key) secret.
2 You encode your credit card details using the public key, and send the
result to the shop.
3 The shop decodes your information using their private key.
This way, it doesn’t matter who gets to know the public key – it can liter-
ally be published – because all it is good for is writing encoded messages that
can only be read by the holder of the matching, top secret, private key.
How can this work? Obviously, there must be a mathematical rela-
tionship between the two keys – so why can’t an attacker work out the
private key from the public one? The best-known method of public key
Pick two prime numbers p, q p = 5, q = 11

Calculate the ‘modulus’ m = pq m = 55
Calculate f = (p − 1)(q − 1) f = 40
Pick the ‘public exponent’ Epub Epub = 7 This can be any number less
than f that doesn’t share any
factors (apart from 1) with f
Public key (Epub , m) (7, 55)
Pick the ‘private exponent’ Epriv Epriv = 23 Pick so that (Epriv × Epub ) − 1
is exactly divisible by f
Private key (Epriv , m) (23,55)
Encoding with public key E pub = 7, m = 55

Message W I B B L E
t 22 8 1 1 11 4
Epub
t mod m 33 2 1 1 11 49
Decoding with private key E priv = 23, m = 55
t 33 2 1 1 11 49
Epriv
t mod m 22 8 1 1 11 4
Message W I B B L E
FIGURE 13.6.2 Trivial illustration of public key encryption.
encryption – RSA – uses the properties of prime numbers. Figure 13.6.2

shows a (trivialised) example of how it works. The parts of the keys 7, 23,
and 55 are all determined by the pair of secret prime numbers 5 and 11, but
in order to calculate the private key (23,55) from the public key (7,55) you’d
need to work out that 55 was the product of 5 and 11. This looks pretty
simple (there aren’t many prime numbers below 55 that it could be) but in
real life, the prime numbers used are so large that the product is hundreds of
digits long – and finding the prime factors of such large numbers is an enor-
mous task even for modern computers.
Note that, apart from using trivially small prime numbers, there are many
flaws in the scheme in Figure 13.6.2. For one thing, encoding text letter-
by-letter like that with the same key makes the whole thing susceptible to
the old high school letter frequency crack. To avoid this, chunks of the
message would be represented as large numbers (imagine ‘220801011104’
for ‘WIBBLE’ – although a base 2 encoding would be more likely) and each
very large number chunk encoded in one go. Combine that with using far
larger prime numbers and you begin to see the other issue with PKE. It uses
a lot of computer power to do arithmetic on very long numbers. Even our

trivial example resulted in numbers too large for Excel to handle. That’s
why Figure 13.6.2 example shows PKE being used to encode just the key-
word WIBBLE from Figure 13.6.1. Real public key encryption is often just
used to securely exchange a key to use with a less time-consuming cypher
for the message.
Technical aside: You may have heard quantum encryption being discussed
as something needed for the day when computers get powerful enough to
crack public key encryption. In essence, the ‘quantum’ part provides a way of
two people securely agreeing on a random number – in a way that is impervi-
ous to eavesdropping – which they then use as the key for a less exotic cypher
to be transmitted by conventional means.
Having a good secret code is one thing, but how can you be sure who you
are exchanging messages with? A useful feature of public key encryption
is that it works the other way round – if you encode something using your
private key, anybody can decode it using the public key. This can be used to
verify the identity of someone you are talking to.
• Alice speaking publicly, announces “My public key is KA”.

• Bob meets someone claiming to be Alice. Bob gives this maybe-Alice a ran-
dom word and asks them to “sign it” (i.e. encode it with their private key).
• Bob then checks the result by decoding it with Alice’s widely known public
key. If that works, it means that it must have been encoded with Alice’s
private key. This is as close to identifying maybe-Alice as the real Alice as
mere mathematics can get. Of course, it all relies on the real Alice keeping
their private key secret.
• This can easily be expanded into a two-way ‘handshake’ whereby Alice
verified Bob’s identity at the same time.
This is the basis of secure “digital signatures”. If, say, you want to be sure
that a website really belongs to your bank, this depends on some trusted third
party whose job is to verify the identity of website operators and vouch for
their public keys. If you want to create a website that shows up as ‘secure’ in
people’s web browsers, then you have to register with a “certificate provider”
who must somehow confirm your identity. You then end up with a certificate
containing your public key that has likewise been signed by the certificate
provider to confirm that it is genuine.
Unfortunately, there is huge demand for these certificates and the process
of issuing them is mostly automated – nobody will come around, check your
birth certificate and meet your grandparents. In most cases all that will hap-
pen is that the certificate provider will confirm your email address and check
that you are able to place a test file on your website. A scheme for ‘extended
validation’ certificates where providers did do more stringent identity checks,
and which displayed the name of the certificate holder in the web browser
address bar has largely failed for various practical, economic, and political
reasons – partly because it still made mistakes and also proved largely inef-
fective at preventing real-world fraud. So, when it comes to web pages, all the
padlock icon really means is that your data is being encrypted, so it is hard
for a third party to intercept. If someone has tricked you into visiting a fake
website, you’re on your own.
Digital signing using public key encryption has other valuable uses,
though. It is a better way of implementing passwords: rather than sending
your secret password to someone to be checked against their records, they
have your public key and use it to send you a ‘challenge’ to sign with your
private key. In systems like this, your password is often just an extra precau-
tion to protect the file containing your private key and neither your private
key nor the password ever need leave your possession.
13.7 How does this fit in a mathematical literacy curriculum?

The chapter is very much about mathematics that students should per-
haps know about without needing to ‘know how’ to implement the details.
Perhaps these topics are best seen as an attempt to promote curiosity (see
Chapter 11) in the way computers actually solve problems and how sub-
jects like sorting and organising data are actually an important part of
modern mathematics. Even if some of the material in this chapter is too
challenging for the typical classroom, wherever subjects like series expan-
sions, logarithms, networks, permutations, or prime numbers do arise in
the curriculum, it is well worth finding time for a brief discussion of their
significance in computing.
Acknowledgements
Figure 13.3.2 is a screenshot from the freeCODAP tool © 2018 The Concord
Consortium – the context comes from the Australian Academy of Science’s
reSolve: Maths by Inquiry Project. See link [13E] below. Figure 13.5.1 is
from the XKCD comic strip by Randall Munroe [13F] and is available under
the Creative Commons Attribution-NonCommercial 2.5 Licence.
References
Knuth, D. E. (1973). The art of computer programming, volume 3: Searching and
sorting. Addison-Wesley.
Press, W. H., Flannery, B. P., Teukolsky, S. A., & Vetterling, W. T. (1986). Numerical
recipes: The art of scientific computing. Cambridge University Press.

purposes:
[13A] Simple Quadratic Equation Solution (Spreadsheet & code)

[13B] Rounding Errors (Spreadsheet & code)
[13C] A More Accurate Spreadsheet Quadratic Solution (Spreadsheet &
code)
[13D] Taylor’s series (Spreadsheet)
https://ltml.mathlit.org/13D
[13E] Monte Carlo Simulations – reSolve & The Concord Consortium
https://resolve.edu.au/monte-carlo-simulations
https://codap.concord.org/releases/latest/static/dg/en/cert/index.
html#shared=50813
[13F] Password Strength – xkcd.com
https://xkcd.com/936/
[13G] Encoding with a secret key (spreadsheet)
https://ltml.mathlit.org/13G
14
REFLECTIONS
Half a century ago one of us set out an agenda for bringing ‘the real world’
into school mathematics (Burkhardt, 1981). Then, together in Nottingham
University’s Shell Centre for Mathematical Education, we began to develop
teaching materials to support this aim (see, for example, Burkhardt et al.,
1980). Since then, each of us has made mathematical literacy an important
part of our work through a series of research-based design and development
projects [14A]. Some of the fruits of these projects have appeared in the vari-
ous chapters of this book.
In writing this book, our aim has been to draw together into a coherent
whole the insights from our contributions and those of others to the teach-
ing of mathematical literacy. We have provided access to a diverse range of
rich lessons that teachers have found to work well in getting their students
to see mathematics as useful for their current and future lives. We have set
these lessons in a framework for mathematics that is firmly context-focused,
integrating the ‘know how’ and the ‘know about’ of mathematical modelling
and data analysis within a critical thinking approach. Experience suggests
that this helps students to develop a productive disposition towards looking
at the world with a mathematician’s lens.
We believe that teachers who work with this approach will find it as use-
ful and enjoyable as others have. In this final chapter we reflect a little more
deeply on some of the key issues that have come up in previous chapters.
14.1 On the context-focused mathematics framework

At the outset of writing this book, we saw that mathematics educators had de-
vised multiple informative frameworks to organise the different components
DOI: 10.4324/9781003303503-15
Reflections 269
of mathematical literacy (or numeracy, quantitative literacy, statistical or

data literacy, depending on local terminology). Individual frameworks high-
lighted different components and linked them in various ways, but there is
considerable agreement between them. They all included consideration of the
context, mathematical and statistical knowledge (concepts, skills and strate-
gies and usually use of tools), along with aspects of productive disposition
(especially confidence) and underpinned by critical thinking about the con-
text and the mathematics. These components seem to describe the task of
teaching for mathematical literacy in schools well.
We wanted this book to show mathematical literacy in practice, illustrated
by situations from real-world contexts that may be encountered by young
adults. All components of mathematical literacy were evident within the set
of examples that we explored. Yet this process highlighted for us the value
of ‘knowing about’ as well as the school focus of ‘knowing how’. In many
cases a mathematically literate person (i.e. one who could make good deci-
sions about a situation informed by quantitative thinking) needed a strong
‘know how’ of the basic mathematics/data literacy/tool use, but a more gen-
eral ‘know about’ across more advanced ideas. By advanced, we mean math-
ematics beyond what is generally taught in the compulsory years of schooling
up to about age 14. ‘Know how’ for advanced mathematics definitely con-
tributes to the mathematical literacy of people working in quantitative pro-
fessions, but for others, ‘know about’ is sufficient.
For this reason, we developed the context-focused mathematics frame-
work (see Figure 1.1.1) for this book to include a wide ranging ‘know how’
component and a wide ranging ‘know about’ component. These two com-
ponents each encompass knowledge, concepts, skills, and processes related
to mathematics, statistics, data science, and digital tools. Some chapters in
familiar areas – on planning and money, for example – have a ‘know how’
emphasis, with plenty of opportunity for modelling or calculating. Others,
in more complex areas like climate change and inequality, are mainly ‘know
about’, designed to support critical thinking about the data and the models
that experts create, their assumptions and consequences. Life-related con-
texts have been the focus throughout.
This raises important questions. Teaching mathematics ‘know how’ for
mathematical literacy is well researched and there is broad agreement on the
special emphases required for mathematical literacy, outlined in Chapter 3
and evident in almost every chapter of the book. But how can we teach the
‘know about’ component well? Could the know about component be written
into curriculum and could it be assessed? There are many questions. Does
‘know about’ accumulate from just a little ‘know how’ exposure (e.g. learn
a little coding to understand how computers do what they are programmed
to do). Perhaps ‘know about’ needs substantial ‘know how’ to see the fun-
damentals of a topic and remember them (e.g. to appreciate the importance
270 Reflections
of functional behaviour). Would including ‘know about’ goals in a teach-

ing program (for example, by writing an essay on how weather is forecast
or the dilemmas raised by inequality) make a lasting difference, or simply
waste time. We know that spending a lot of time on routine practice (e.g.
of arithmetic or algebraic skills) does not result in ‘know about’ (or, indeed,
long-term ‘know how’) but is it possible to produce ‘know about’ without se-
rious attention to some ‘know how’ skills? And the answers may be distinctly
different for the many different things a mathematically literate person may
need to know about. We have shown examples of how this has been done.
Another important decision in the book is to present theory-driven math-
ematical modelling and data-driven investigation as two approaches to an-
swering a real-world concern. Some of the chapters lean to the theory-driven
side, and others are data-driven, but all of the chapters show how intertwined
these approaches can be in practice. A surprising realisation for us was to ob-
serve the great dependence of society today on measures that are, at heart,
mathematical models of an abstract concept. These are used to compare or
to track changes of things as diverse as the health of a national economy,
how well washing machines perform, obesity, and fairness of society. Each
depends on creating a mathematical model that identifies significant variables
(and devising ways of measuring them) and then identifying the relationships
between the set of variables that might give a useful measure. In this way, we
see that a great deal of data is actually the product of mathematical model-
ling, further intertwining the theory-driven and data-driven approaches.
14.2 On the roles of technology

Technological advances have always been used to facilitate ‘doing math-
ematics’ – from making marks in the sand, via using an abacus to the latest
computers – silicon to silicon in four millennia! New advances in these
technologies have often raised concerns about the loss of familiar skills. In
a speech on administrative reform Charles Dickens (1855) referred to the
use of notched splints of elm called ‘tallies’ to keep the Exchequer accounts,
which continued until 1826 even after an enquiry “by some revolutionary
spirit” asking whether, “pens, ink, and paper, slates and pencils, being in
existence, this obstinate adherence to an obsolete custom ought to be con-
tinued” – a situation not dissimilar to the resistance to the introduction of
calculators in school mathematics classes a couple of centuries later, which
has continued to the present day. In retrospect, each technological advance
enabled a person to do more with less effort. Tables of logarithms were a
prime example, enabling more precise calculation for navigation and many
other purposes. In the last half century, we have stepped up from formula
books to calculators of increasingly sophisticated types, and to computer
software from GeoGebra to Mathematica and Wolfram Alpha. Each new
Reflections 271
tool has shifted the skills and knowledge required by someone to use them
effectively to apply mathematics in understanding the real world. Recent
‘large language models’ (LLMs) like ChatGPT have the potential, used intel-
ligently, to significantly increase the power of the individual as a mathemati-
cally literate person – although the initial concern from education seems to
focus on their potential as a way of cheating on essays and dissertations.
But first on the basics. The school curriculum inevitably moves at a slower
pace than the technology that is widely available to the population at a rea-
sonable cost. But spreadsheets, for example, have been standard tools on
home and work computers for about 30 years. As the chapters of this book
illustrate, they are extremely useful to support a wide range of calculation
and exploration. Making spreadsheet use a normal part of school mathemat-
ics is surely long overdue. There are, of course, other examples.
Looking forward, the current capabilities of LLMs could be indicating a
major shift in those skills and knowledge may be required. If you can frame
a problem in mathematical language (a formula, for example), then you can
probably get a solution from a spreadsheet, or even ask Wolfram Alpha. But
with LLMs, we may be able to ask open-ended questions in (non-mathemat-
ical) ‘natural language’ and get
a direction on how to understand the problem,

b mathematical solutions, and
c explanations on how to interpret and understand the answer.
For example, backpackers are using ChatGPT to plan their trips between
one country and the next, giving travel arrangements within a budget, sites
to see on route, and so on. People are using ChatGPT to help plan a party,
including organising a schedule, managing guest invitations, budgeting for
food/drink, and so on. It is used to get estimates of physical quantities: “es-
timate the force on the rope when a rock climber falls from 1 m above the
bolt”, “estimate how far away the horizon you can see is when you are on a
boat surrounded by ocean”. You can give ChatGPT a spreadsheet file and ask
it to produce a basic data analysis.
These tools may make a difference to being a teacher – “Write me 10
questions on factorising quadratics”; “Find me a problem suitable for a one
week assignment for year 9 involving similar triangles”. Better check them,
of course!
These new capabilities pose new potentially great opportunities for people
to benefit from applying mathematics for the betterment of their life and their
decisions, but they also surely pose new challenges and demand new skills of
the kind we have foregrounded in this book. How do you interface effectively
with LLM services – the next generation of the challenge of learning how
to ‘Google’ effectively? How do you cross-check what an LLM tells you?
272 Reflections
What ways are there to verify answers? Since LLMs sometimes give wrong
answers, corroboration is essential. The list of new challenges is long.
Beyond this are the multiple varieties of potential damage that LLMs can
do to society in plagiarism, malign messaging or just misinformation. This
has already been exemplified by much less powerful and plausible social me-
dia. How this can be prevented – or, more likely, mitigated?
Critical thinking by mathematically and socially literate people surely has
a role to play in the effective and positive use of LLMs – and is one thing that
these systems cannot currently do for us. If students can easily ‘cheat’ by get-
ting LLMs to complete their assignments, then the problem may be that those
assignments are not requiring sufficient critical thought.
We end this forward look optimistically, with the last line of George Ber-
nard Shaw’s ‘metabiological pentateuch’ Back to Methuselah (Shaw, 1921).
And for what may be beyond, the eyesight of Lilith is too short. It is
enough that there is a beyond.
14.3 Developing professional expertise

It is clear that the variety of classroom activities and the range of contexts
that we have described above will ask that many teachers of mathematics
move beyond the comfort zone of their current practice. The sense-making
approach, summarised in TRU and the context-focused mathematics frame-
work, implies a broader range of concerns than is traditional in school math-
ematics – ‘know about’ as well as ‘know how’, with critical thinking about
both the context and the mathematics and, above all, a focus on what the
student is doing and the affordances the activities offer.
Some teachers welcome the opportunity to make mathematics more useful
for their students; for them we have shown a wide range of both classroom
and professional development activities that offer effective support for broad-
ening the range of their expertise. Others are unenthusiastic – their job is
demanding enough already. Yet there are cogent reasons for all teachers and
teacher educators to explore this important domain with their students. For
most students, mathematical activities focused on life-related contexts give
added meaning to the mathematics, and motivation for developing their con-
ceptual understanding and procedural skills – and they enjoy the activities.
For the individual, there are benefits in their personal, work, and civic lives.
For society the primary goal of school mathematics, and the justification for
its large time allocation, is to produce appropriately mathematically literate
adults for all levels of the workforce.
We hope we have encouraged all teachers of mathematics to move, how-
ever gradually, to include mathematical literacy as an aspect of many of their
lessons and the focus of some.
Reflections 273
And finally…
We have learned a lot in writing this book – and enjoyed the process. We
hope that you, the reader who has got this far, have found this exploration
with mathematics of some aspects of the world we live in enjoyable and
informative – and that your students will too.
References
Burkhardt, H. (1981). The real world and mathematics. Blackie-Birkhauser.
Burkhardt, H., Treilibs, V., Stacey, K., & Swan, M. (1980). Beginning to tackle real
problems. Shell Centre Publications.
Dickens, C. (1855). “Administrative Reform” (27 June 1855), Theatre Royal, Drury
Lane Speeches literary and social by Charles Dickens (1870) pp. 133–134.
Shaw, B. (1921). Back to Methuselah: A metabiological pentateuch (Vol. 16). London
Constable 1922.

To visit any of these links, scan this QR code or visit ltml.math-
lit.org – append the link code to go directly to the entry – for
example, ltml.mathlit.org/14A
purposes:
[14A] Past projects by the authors

INDEX
3X teaching 55 Asking for permission (task) 37

4 Ps (in marketing) 145 Assessing Mathematical Literacy 33
assumptions 34, 44, 46
accounting 172 authenticity 43, 50
accuracy 242 average 196
active learning 68 average income/wage 159–161, 164,
active modelling 16, 53 184; versus median 159
advertisements 141
advertising 145, 148 Babbage, C. 207, 248
agency 52, 54, 57 Balanced Assessment Project 227
Airplane Turn-round (task) 11, 25, 125, bank account 176
226 bank statement 174, 187
algebra 59, 133, 196, 240 Barwell, R. 121
algorithm 253, 257–258; game playing base risk 78, 80, 95, 98
253; order/big O notation 249; basic income 166
search 249; social media 157; basket of goods 232
sorting 248 bathtub curve 151, 187
all models are wrong but some are Be a Paper Engineer 17, 40
useful 16, 72 bias 142
analytic approach 9 bias by balance 155
angles 41 bias in artificial intelligence 258
animation 193, 197, 200 Bicycle cornering (task) 199
app 102, 199 binary numbers and fractions
applications (of mathematics) 16 243–244
apprentice task 44 biodiversity 67–68
appropriate accuracy 60 biofuels 120
appropriate precision 81 biology 67–68
argument 70, 153 Black, P. 234
arithmetic 26 Black, P. and Wiliam, D. 45
arithmetic mean 234 black death 94
artificial intelligence 89, 136, 192, Blum, W. and Niss, M. 15
258, 271 Body Mass Index (BMI) 226
Index 275
bonds 185 comparing different representations

boundaries 132 56
Bowland Maths 19, 39, 55, 72, 81, 87, comparing prices 141
120, 128, 207, 216, 222 comparing products 141
Bowland Maths professional comparing student work 45
development 72 complexity 26
box 72 computer see information technology
Brief History of Equality, A 160 computer algebra system 192, 245
Brousseau 55 computer errors 206
bubble chart 63 concepts and skills for mathematical
bubble sort 248 literacy 57
budget 173 conceptual demand 33
bullshit 157 conceptual model 14
Burkhardt, H. 45, 60, 221, 234, 268 conditional probability 99
butterfly effect 24 confirmation bias 156–157
constraints 131, 133
calculator 21, 23, 50, 192, 270; consultant (role playing) 54
graphing 134, 196 consumer decisions 146
calculus 116, 240 consumer magazines 141
capital 175, 178, 184 consumer price(s) index see
carbon dating 110 content 52–53
carbon dioxide 107–111, 117, 119–120, context 8–9, 12, 14, 43, 54, 64,
226 69; border or tapestry 60;
carbon offsetting 114 complexity & familiarity 9
cash flow 175 context-focused mathematics framework
catastrophe theory 118 3–4, 6, 31, 42, 51, 52, 213, 241,
causes of death 19 268–269, 272
chains of reasoning 56, 102, 233 Continent area (task) 35
chance see probability continuous mathematics 240
ChatGPT see artificial intelligence controversial and sensitive issues 69
checking tactics 56 Cornering (task) 199
Choice (publication) 143 correlation 166, 230; versus causation
cholera 135 166
CIA 89, 204 COVID-19 13, 16, 77, 80, 90, 92–94,
circles 41 96, 98, 174, 232; modelling 97;
class discussion 69 mutation 95; outcomes with age
classroom challenges see mathematics 95; vaccination 98
assessment project CPI 234
classroom contract 55–56 credit 176
climate change 69, 106–123; credit card 174, 176, 183
anthropogenic 107, 114; politics credit rating 176
112, 114 credit versus debt 176
climate change versus environment 119 critical enquiry 12
Clothes airer (task) 62 critical path 127–128
CO2 see carbon dioxide critical thinking 4, 7–8, 12, 39, 50,
coding see programming 53–54, 68–69, 71, 86, 121, 140,
cognitive demand 52–53 159, 269
Coin Counting (task) 208, 222 critical thinking about mathematical
Cola Pool (task) 222 models 70
command of language 12 cross curriculum see interdisciplinary
Common Core State Standards 45 work
common issues 45 cryptography see encryption
common sense 50 CSIRO 204
276 Index
curiosity 8, 213–225; concern 218; doubling time 92, 179–180; rule of

developing in the classroom thumb 181
220; information seeking 220; drama (in lesson design) 200, 221–222
knowledge 219; novelty 220; drowning 86
opportunity 218; perceptual and Drucker, P. 226
epistemic 214; stimulatng 217; dynamic geometry see interactive
success 219; surprise 217 geometry
curriculum 3, 23, 31, 50, 63
curriculum time 23, 49 earnings see income
cypher 262; substitution 262 earthquake 88
ecology 68, 202
data 7–8, 53, 58, 85, 269; big 138; economic forecasting 152
collection 68, 135; demographic economics 161
157, 204; displays 58–59, 63; ecosystem 119, 121
economic 204; handling 194, Eisner, E. W. 223
202; literacy 12, 18–19, 76, 85, encryption 262; public key 263–265
88–89; realistic 202; reliability energy production 238
88–89, 156; science 134, 136– engagement 64
137; selection of 153; sources enlarging 66
89, 204; storing and organising environment 119–120
248; use, or abuse, of 140; epidemic 90, 92, 94, 135; cholera 135
visualisation 89 epidemic curve 90–91
data-driven approach 9, 12, 18, 31, 34, Equality Trust 164
91, 136–137, 270 equality of wealth and income 159–169
database 55, 194, 248, 251–252; keys equation, approximate solution 193;
and indexes 249, 252; relational difference 181; quadratic see
249–253 quadratic equation
death rates 92 equity 52, 54
debit card 174, 176 estimation 39, 57, 59–60, 146, 207;
debt 174, 180, 183 approximate calculation 61,
decathlon 144 207; checking of calculations 61;
decay 178 plausible 207
deliberative thinking 76 ethics 69
descriptive 9 evaluate 11, 14–15, 18–19, 32
design a board game 40, 130 evaluate and improve 31, 34, 36, 38–39,
Design a measure 230 43
design and make 31, 37, 43 everyday life 1, 3, 7, 16, 25, 34, 38, 50,
design and make a party invitation 17, 170, 208, 215, 240, 262
19 examples, importance of 2, 25
Desmos 134, 200–201 excess deaths 99
Dewey, J. 215 exercises 44, 56
diesel engines 119 expenditure 170, 173–175
Dietiker, L. 222–223 expert task 44
differential equation 116 exploratory data analysis 20
difficulty (of a mathematics task) 44 explore 43
digital signatures 265 explore and discover 31, 42
dimensions 33 exponential 9, 97, 116, 178
disasters 77 exponential decay 59
discounting 188 exponential growth 90, 92, 183; and
discrete mathematics 240 money 179
discretionary spending 141 extended chains of reasoning 34
discussion 46, 54; and reflection 34 extended warranties 1, 151, 187
dispositions 66 external validity 89
Index 277
fairness 39 Goos et al. 12, 65

false negative 99 grains of wheat/rice on a chessboard 44,
false positive 99–100, 102 179
fatal accident 18 graph 26, 37, 193, 197, 203
feedback 107, 111, 115 graph sketching 201
Fermi estimate 207 graphing calculator 134
Ferris wheel (task) 32–33 Graphing Stories 200–201
few year gap 53 greenhouse effect 111–112
Feynman, R. 56 greenhouse gases 111–112
financial adviser 178, 185 gross domestic product see GDP
financial literacy 2, 171 group work 41, 46, 69, 78, 115;
finance, values and goals 178 presentation 57
finding reliable data 88 growth 163, 178, 240
five components of mathematical guess–check–improve 23
literacy 7
floating point 243, 245 headlines 153
Forman, S. L. 44 Height and arm span (task) 196–197
formative assessment 41, 45, 52, 54, 61 herd immunity 90
formulate 10, 14–15, 18, 31 heuristics 253–254
fossil fuels 112–114, 120 high stakes assessments 233
framework see context-focused house prices 184
mathematics framework How risky is life? (materials) 19, 76, 82
Freudenthal Institute 18 Howard, R. A. 78
function 133, 136; exponential 59–60; Hoyles, C. 192
graphical representation 196; humanities 69
inverse proportion 59; linear Hurdles race (task) 19–20
59, 196–197; potential 118;
predictive 137; series expansions ice cores 108, 110
247; symbolic representation ice cover 116
196; trigonometric 246; utility identity 52, 54
132–133, 143–144 identity verification 265
image recognition 257
gallery walk 57 image sizes and resolutions 209–210
games (algorithms for) 253 immunity 90, 92
Gantt chart 126–128 income 159–161, 164, 170, 172–175,
GapMinder 12, 63, 70, 89, 156–157, 177, 235
204 IMPACT 4
garbage in, garbage out 207 independent routes 56
Gattegno, C. 215 index of health and social problems 165
GDP 63, 72, 153–154, 161, 238 inequality of wealth and income 159–
GDP growth 154 169, 235
Geiger et. al. 6 infection 90
gender parity index 205 infection–fatality ratio 95
GeoGebra 134, 199, 270 inflation 181–183, 188, 227, 234;
geography 68, 202 basket of goods 182; see also
geometric mean 234 CPI, RPI
geometric theorems 41 influencer 142
geometry 41, 58, 240; interactive 199 influenza 94
getting meaning out of data 26, 37 information entropy 259
Gigabytes 147 information technology 21, 50, 191–
Gini Coefficient 235 212, 240–267
global warming potential (GWP) 111 information technology and the need for
Goodhart’s law 231 mathematical skills 192
278 Index
information technology as a teaching logarithm 246, 270; in music 232

tool 193 logarithmic scale 82, 96–97, 181, 232
inheritance 166 logistics 134–136
insurance 187 LOGO (programming language) 210,
interactive geometry 199–200, 202 215
interdisciplinary work 64 Lorenz curve 236
interest 174, 176, 178, 182, 227; loss leader 145
compound 178, 180
interest rates 9, 180 183, 188 machine learning 138, 256, 258; see also
International Society for Design and artificial intelligence
Development in Education Making and Selling a Magazine (task)
(ISDDE) 37 193–195
interpret 10, 14–15, 18–19, 36, 43 Making matchsticks (task) 39
interpret and translate 31 mantissa 244
investigation 64 MAP see Mathematics Assessment
investment 180, 185 Project
Investment Game (task) 186 maps 58
isometric drawing 65 market and brand analysis 145
isotope 110 marketing 141
mathematica 270
Kahneman, D. 26; see also slow mathematical literacy 1, 5–6; meaning
thinking and interpretation 24–25
kilogram 237 mathematical literacy approach (by
King Camp Gillette 150 topic) 57–59
know about 4, 8–9, 53, 70, 77, 87, 106, mathematical model see modelling
125, 159, 240, 269 Mathematics Assessment Project (MAP)
know how 7, 9, 13, 42, 53, 77, 125, 38, 45, 100, 120, 133, 230
269 maximising profit 194
know how versus know about 9 Maximising profits: Selling boomerangs
(task) 133
language (literacy) 2 maximum 119
Language of Functions and Graphs 20, measure 4, 226–238, 270; of appetite
36–37 for risk 185; of biodiversity
large language models see artificial 67; of compact-ness 229; of
intelligence correlation 230; of inflation 227,
learning goal 3 234; of inequality 227, 235; of
least squares 230 likelihood 80; of obesity 226; of
levels of expertise 44 password strength 259; of risk
Licorice Allsorts (task) 65 78; of squareness 228
life expectancy 63 measurement 57, 66
life-related, 1–4, 7–9, 11, 15, 26, 30, Mechanical Linkages and Deductive
44–45, 50–54, 231, 269 Geometry (materials) 40–42
life tables 79 media, news 155
lifetime earnings 177 median income/wage 153, 159, 184
linear programming 133–134 medical testing 99
links and references 4 metastable equilibrium 119
literacy 24; data 12, 18–19, 76, 85, meteorite 77
88–89; financial 2, 171; health method of differences 248
2; information technology 2; rate metre 237
204–205; scientific 2; statistical Meyer, D. 200, 221–222
2, 18, 269; techno-mathematical micromort 78–82
13, 192 minimum 119
loan 9, 177 misinformation 157; see also spin
Index 279
misunderstandings 52 optional extras 150

mobile phone 21, 146 orders of magnitude 208
model and explain 31–32, 40, 43 Organisation for Economic Co-
modelling 8, 10–11, 13–14, 51, 53, operation and Development
270; apps 71; climate 116; with (OECD) see PISA
computers 193; COVID-19 organising events 125
97; cycle/process 10–16, 31, ownership 52, 54
193–194; epidemics 90–91; Oxygen (task) 198
feedback 116; habitat following
a fire 202; ice cream sales 136; pandemic 13, 77, 80, 90, 92–93, 174
interactive 199; phases of Papert, S. 215
10–11, 14, 31, 34; probabilistic parallel calculation 56
136; route planning 71; social participatory socialism 166
constructs 166; social media 71; password 258, 260–261, 266
with spreadsheets 194; statistical password entropy 232, 259
137; tipping points 117; weather password manager 262
prediction 72; what if? 194 password versus passphrase 261
money 170–190; borrowing 183; buying paying monthly 150
power 227; value of 181 Pead, D. 210, 221
Monte Carlo methods 255 pedagogy 52; active learning 68; for
Moore’s law 209 planning skills 128; qualitative
mortality rates 95 enlargement of the range of
motivation 43 learning goals 72
Mr. Micawber 170 pen and paper mathematical skills 50
Muddying the Waters (task) 120 percents 58
Munroe, R. 261 permafrost 117
personal finance 170–188
national income 153 Pert diagram 126–127
national statistical agencies 89, 155 phases of modelling 10–11, 14, 31, 34
navigation 58, 71 phenomena 4–5, 8, 12, 31–32, 57, 70,
needs versus wants 141, 176 119, 214, 216, 221, 226, 237
neural network 256–258 photosynthesis 198
news and current affairs 152, 155 PIAAC 6
Newton’s method 247 Pickett, K. 164, 166
nomenclature 24 picture graph confusion 37
non-routine problems and tasks 45, 53, Piketty, T. 160
233 PISA 3, 13, 24–26, 32–33, 35, 43, 47,
noughts and crosses 253–254 88, 171, 233, 235
novice task 44 PISA mathematics framework 5
NRICH 144 Plan a Trip (materials) 40, 130
nuclear power 120 plan and organise 31, 34, 40, 43, 125
number sense 61 planning 124–139, 230
numeracy 2, 24, 269; functional 68 point of inflexion 119
Numeracy Across the Curriculum Poisson distribution 87
65–66, 202 pollution 119–120
Numeracy through Problem Solving 17, population 63, 90, 160
40, 130–131 population density 229
numerical methods 23, 116 poster 57
powers of 10 57
odds see probability precision 44, 58
operational research 135, 138 present value of future income 188
optimisation 133 presentation 57, 69
optimising within constraints 131–132 price setting 145
280 Index
primary key 252 relational database 249–253

primary sources 155 relationships 34
prime numbers 264 reliability 56
probability 58, 67, 77, 80; conditional renewable energy 120
99 report 11, 57
Problems With Patterns and Numbers representing 46
216 Representing Probabilities: Medical
Produce a Quiz Show 40, 130 Testing (materials) 100
product 57 reSolve: mathematics by inquiry 15,
product comparison websites 141 40–42, 47, 82, 86–87, 199, 216,
productive disposition 8, 12, 54, 68, 76, 255
213, 269 retail price(s) index see RPI
productive struggle 52 reviews (or products) 141–142
professional development 39, 51, 72, Revolving Door (task) 34–35
128; sandwich model 72 rice 62; on a chess board 44, 179;
Programme for International Student cooking 71
Assessment see PISA Richter scale 232
Programme for the International risk 58, 76–103, 122; appetite for 185;
Assessment of Adult as a function of age 79; base
Competencies see PIAAC 78, 80, 95, 98; of death 78–79,
programming 50, 210, 241–242, 269 83–84, 95; disasters 77; driving
progress (in learning mathematics) 30 78; of dying 19; financial 185;
proportion 39 gender differences 84; health 78;
proportional reasoning 16, 26, 32–33, impression versus reality 82; in
58 financial matters 171; likelihood
public key encryption 263, 266 versus hazard 77–78; road
purchasing power parity 161 accidents 86; school shootings
80; terrorism 82; versus reward
quadratic 242 185; visualisation 82
quadratic equation 23, 242, 245–246; road safety/accidents 55, 78, 86
formula 23, 242 robust skills 51
quadrilaterals 41 role shifting 55
quantitative literacy 2, 269 Rosling, H. 12, 20
Questioning and Reasoning (materials) rounding errors 244–245
39 route planning 71
RPI (inflation) 234
R number (for infection) 90, 95, 98
radioactivity 110, 120 sample responses 46
random variation 86–87 Sauce (task) 13, 16, 32
rate 116, 175 savings 180, 185
rate of breakdowns 151 Sawyer, J. 107
rate of change 59 scale factor 33
rates 58, 155 scaling problem 32
ratio 58, 66 scam 150
real-life 8, 41, 61, 86, 264; see also scatter graphs 197
life-related Schoenfeld, A. 52, 216
real-world 2, 10, 33, 53, 196, 268; see science 202
also everyday life, life-related scientific notation 243
Realistic Mathematics Education (RME) Scratch (programming language)
18 210
redistributive tax 161 sense making 56
Reducing Road Accidents (materials) sensitive issues 69
55, 57, 222 shares 186
Index 281
Sharing gasoline costs (task) 38 table tennis tournament 34, 125,

Shaughnessy, J. M. 58 128–129, 216
Shell Centre For Mathematical tackling unstructured problems in the
Education 17, 36–37, 40, 268 classroom 128
shooting at school 1 taking the context seriously 7, 10–11,
SI units and prefixes 209, 237–238 44, 53, 55, 59, 140
significand see mantissa tariffs 146
Similar Triangles (Ironing Table - task) tasks 30–48; 5–10 minute 34; attributes
40 of 42; complexity 23, 44;
Simpson’s diversity index 67–68, 72 difficulty 44; familiarity 44;
Simulating Random Variations (activity) lesson length 38; multi lesson 40;
86 occupational 43; open 34, 45,
simulation 24, 86, 193, 197–198 128, 130; openness 44; personal
skills 7, 25, 193 43; scientific 43; societal 43;
skills deficit 192 structured 34, 128; types 42;
slow thinking 26, 76, 124 types of 31, 43; variety of 30
Smart::test (materials) 61–62 tax 163, 168; progressive 167
smartphone 146 Taylor’s series 181, 247
Snow, J. 135 teacher educators 1
social constructivism 215 teaching 49–75
social media 142, 155, 157; algorithms Teaching for Robust Understanding
157; misinformation 157 see TRU
social sciences 69 technical skill 60
society 69 technical skills 26
solve 10, 14–15, 18 techno-mathematical skills/literacies 13,
sophisticated use of elementary 192
mathematics 25–26, 44, 54 technology 7–8, 21, 26, 50, 53, 58, 270
sorting 248 The Spirit Level: Why More Equal
Space Flight (task) 32 Societies Almost Always Do
speed time graph 37 Better 164, 166, 235
Spiegelhalter, D. 99, 103 theory-driven approach 9–11, 13, 31,
spin 140–158 91, 270
spreadsheet 15, 23, 50, 58, 60, 90, 100, Thom, R. 118
127, 136–137, 148, 172, 179, three act math 222; see also drama
180, 183, 192, 194, 196–197, three dimensions 58, 66
240, 244–247, 253, 271 tic tac toe see noughts and crosses
square root 246 tipping points 107, 115, 117–118
Stacey, K. 15, 33, 43, 61, 202, 216 tools 50; curve fitting 201; digital tools
standard applications 16–17, 53 50; mathematical 50; online 134;
standard models versus active modelling visualisation 199, 204
16 total cost of ownership 147
standards 237 transmissibility (infection) 91
statistical literacy 18, 269 trapezoid rule 247
statistics 58, 63 tree diagram 100
Steen, L. 25, 44, 54 trendline 196
Stillman, G. 60 trial and improvement 193
student teachers 2 triangles 41
students 2 TRU 52–53, 56, 215, 222–223, 272;
subscription 141, 146, 150, 176 dimensions 52
summary statistics 197 Turner, R. 15, 33
supply chain 144, 146 Turning into a parking space (task)
surprise 223 200
Swan, M. 37, 40, 82, 221 twenty first century skills 51
282 Index
UK Office of National Statistics 92 Watson-Watt, Sir R. 135

UNESCO 89, 204 wealth 159–161, 164, 235;
unintended consequences 231 see also inequality of wealth
unit trusts 185 and income
units 57–58, 63, 155, 175, 208, 237, weather forecasting 24, 72
259 weighting variables 132
Upgrade? (task) 34 what you test is what you get
utility function 132–133, 143–144 144, 233
Which Car? (task) 34–36, 132
vaccine 98 Which Sport? (task) 37
values 69 wholesale 145
variables 9, 15, 18, 34, 37, 44, 63, Wilkinson, R. 164, 166, 235
89, 132, 136; choice of 153; Will It Hit the Hoop? (task) 200
independent and dependent 59 Wolfram Alpha 270–271
Venn diagram 103 word problems 59
versus correlation 166 word processor 192
video (as a stimulus) 193, 200, 202 Worldview upgrader see GapMinder
Visicalc 194 working with others see group
visual communication design 65 work
visualisation 157, 197, 199; tools 204 work–life balance 178
World Bank 89, 204
wages, average 159–161, 164, 184;
versus median 159 You Reckon? (materials) 120, 208
washing machines 143

Sanet - ST 1032301171

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

Sanet - ST 1032301171

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Sanet - ST 1032301171

Uploaded by

Copyright:

Available Formats

LEARNING AND TEACHING

• Describing a framework for a broader vision of mathematical literacy –

Firmly grounded by practical applications for the classroom and

Kaye Stacey is an internationally renowned researcher, teacher educator, and

All three authors have received awards for contributions to educational

The Learning and Teaching of Algebra

The Learning and Teaching of Geometry in Secondary Schools

The Learning and Teaching of Mathematical Modelling

The Learning and Teaching of Number

The Learning and Teaching of Calculus

The Learning and Teaching of Statistics and Probability

Hugh Burkhardt, Daniel Pead,

ISBN: 978-1-032-30116-7 (hbk)

1 What Is Mathematical Literacy? 5

2 The Power of Tasks 30

3 Teaching for Mathematical Literacy 49

4 How Risky Is Life? 76

5 Climate Change: The Science, the Mathematics, and

6 Planning for Good Things in Life 124

7 Looking Past the ‘Spin’ 140

8 Equality and Inequality 159

9 Your Money in Your Life 170

10 Computers in Teaching for Mathematical Literacy 191

11 The Importance of Curiosity 213

12 Designing Measures 226

13 Mathematics for Information Technology 240

IMPACT, an acronym for lnterweaving Mathematics Pedagogy and Content

My dad is buying a new dishwasher. Should he buy the extended warranty

• teacher educators, working in both initial teacher preparation and profes-

• student teachers, especially those preparing to teach teenagers, and the

The aim is to be readable by student teachers coming from a wide range

• a descriptive and analytic text for reading about mathematical literacy –

The form of support we have devised for these objectives is a combination

Structure of the book

The underlying messages in each of these six chapters on teaching for

Links and references

Mathematical literacy is the ability to use one’s mathematical capabilities to

Mathematical literacy is an individual’s capacity to reason mathematically

We believe this broad description summarises the various ways in which

1.1 A context-focused mathematics framework

FIGURE 1.1.1 The context-focused mathematics framework.

The five components of mathematical literacy

Taking the context seriously

Knowing how to use a range of mathematical,

Knowing about modelling, data, and technology

Taking a ‘critical thinking’ approach

Demonstrating a productive disposition

1.2 ‘Know how’ versus ‘Know about’

1.3 Complementary approaches – ‘theory-driven’

sometimes starting with ideas about relationships, sometimes with data of

Theory-driven approach – modelling a situation

FIGURE 1.3.1 The modelling process.

What is the shortest time needed to do all these jobs?

FIGURE 1.3.2 A task demonstrating taking the context seriously.

Critical thinking and critical enquiry

1.4 Modelling real situations with mathematics

You are making your own batch of dressing for salads.

1.3 Complementary approaches – ‘theory-driven’

FIGURE 2.4.2 An explore and discover lesson involving movement of everyday