Elementary Logic as a Tool in Proving Mathematical Statements
 

Bruce Matthew May

A thesis submitted in partial fulfillment of the requirements for the degree of

Magister Scientiae in the Department of Mathematics, University of the Western Cape.

Supervisor: Prof. R.L. Fray

November 2008
  

Romans 11:36
“For of Him, and through Him, and for Him are all things: to Him be glory
forever.”
  
Table of Contents
 
Page
 
Table of Contents i
Abstract   vii
Declaration viii
Acknowledgements ix
List of Figures x
List of Tables xi

Contents

Chapter 1 Introduction

1.1 The influence of Mathematics on Economic Development 1

1.2 Background to the study 1
1.3 The Aim of the study 12
1.4 The Main Research Question 13

Chapter 2 Literature Review

2.1 Introduction 14
2.2 Effect of logic on proving at school level 14
2.3 Effect of logic on proving at tertiary level 16
2.4 Literature review summary 20

Chapter 3 Theoretical Framework

3.1 Introduction 21
3.2 Is proof important in mathematics and if so why? 21
3.3 Origins of proof 22
3.4 Functions of proof 24
3.5 Proving in South African schools 26
3.6 The distinction between a priori and a posteriori knowledge 30
3.7 Cognitive processes involved in learning proof 31
3.8 Physiology of learning 33

i
Chapter 4 The influence of Emotion,  Confidence, Experience and
Practice on the learning  process

4.1 Introduction   35
4.2 The influence of emotion 35
4.3 The influence of Confidence   37
4.4 The influence of Experience 38
4.5 The influence of Practice 39
4.6 Negotiating the learning process 39

Chapter 5 Research Methodology

5.1 Introduction 41
5.2 Learning theory 42
5.3 Teaching Methodology 42
5.4 The logic component of the course 42
5.4.1 Compound Statements 43
5.4.1.1 Truth tables 43
5.4.1.2 Logical equivalence of statements 45
5.4.2 Conditional Statements 46
5.4.2.1 Contrapositive, Converse and Inverse of a conditional
statement 46
5.4.2.2 Biconditional 46
5.4.2.3 Necessary and sufficient conditions 47
5.4.3 Valid and invalid argument forms 47
5.4.3.1 Rules of inference 48
5.4.3.2 Fallacies 49
5.4.3.3 Examples based on argument forms 50
5.4.3.4 Applying the Contradiction Rule 51
5.4.3.5 Marsha’s solution: 52
5.4.3.6 Marsha’s corrected answer 53
5.4.4 Quantified statements 53
5.4.4.1 Contrapositive, Converse and Inverse of Universal
Conditional statements 55
5.4.4.2 Universal Instantiation, Universal Modus
Ponens and Universal Modus Tollens 56
5.4.4.3 Valid and invalid arguments of Quantified statements 57
5.5 The proof component of the course 63
5.5.1 Set theory 64
5.5.1.1 Definitions and Language of set theory 64
5.5.1.2 Set Equality 64
5.5.1.3 Operations on sets 64
5.5.1.4 Element method for proving that one set is a subset
of another 65
5.5.1.5 Element (basic) method for proving that sets are equal 65

ii
 5.5.1.6  
Proving by Division into Cases 65
5.5.1.7 Empty set 65
5.5.1.8 Disjoint sets   65
5.5.1.9 Partition of sets 66
5.5.1.10 Power Sets   66
5.5.1.11 Examples of proofs of set identities and set
 
inclusions where cues on logic are given 67
5.5.2 Elementary Number Theory 70
5.5.3 Methods of Proof 72
5.5.3.1 The method of direct proof 72
5.5.3.2 Disproof by counterexample 77
5.5.3.3 Method of contradiction 78
5.5.3.4 Method of proof by contraposition 81
5.5.3.5 Connection between proof by contradiction
and proof by contraposition 82
5.5.4 Mathematical Induction 84
5.5.4.1 Using mathematical induction to prove
a divisibility property 86
5.5.4.2 Using mathematical induction to prove an
Inequality 86
5.5.4.3 Strong Mathematical Induction 87
5.5.4.4 Teaching to connect ordinary mathematical
induction to strong induction 87
5.5.4.5 Principle of Strong Mathematical Induction 89
5.5.4.6 Proving a property of a sequence
using strong mathematical induction 90

Chapter 6 Results of the study [presentation and discussion]

6.1 Introduction 95
6.2 Logic puzzles 95
6.2.1 Puzzle I (pre – test) 95
6.2.1.1 Analysis of assessment of puzzle I of Experimental Group 96
6.2.1.2 Analysis of assessment of puzzle I of Control Group 97
6.2.1.3 Conclusions based on puzzle I 97
6.2.2 Puzzle III (post test) 98
6.2.2.1 Analysis of assessment of puzzle III (post – test) of
experimental group 99
6.2.2.2 Analysis of assessment of puzzle III of control group 100
6.2.3 Comparison of answers of puzzle I and puzzle III 101
6.2.4 Conclusions based on the analysis of puzzle I and puzzle III 103
6.3 Puzzles on knights and knaves 104
6.3.1 Knights and knaves (pre-test) 104
6.3.1.1 Analysis of arguments of students 105

iii
 6.3.2 Knights and knaves (post - test)  107
6.3.2.1 Analysis of arguments of students 107
6.3.3 Conclusions based on the analysis   of knights and knaves puzzles 109
6.4 Arguments with Quantified statements 111
6.4.1 Pre-test   111
6.4.1.1 Analysis of student answers of Quantified Statements 111
 
6.4.2 Post-test quantified statements 113
6.4.2.1 Analysis of student answers of Quantified Statements 113
6.4.3 Arguments with Quantified statements (forming conclusions) 114
6.4.3.1 Puzzle I (pre – test) 114
6.4.3.2 Analysis of student answers 115
6.4.3.3 Puzzle II (post – test) 116
6.4.3.4 Analysis of student answers 117
6.4.4 Conclusions based on the results of Quantified Statements: 117
6.5 Proofs 117
6.5.1 Set Theory 117
6.5.1.1 Pre-test 117
6.5.1.2 Analysis of student answers 118
6.5.1.3 Set theory (post-test) 121
6.5.1.4 Analysis of student answers 122
6.5.1.5 Conclusions based on the results of set theory 123
6.5.2 Method of direct proof and divisibility 124
6.5.2.1 Pre-test 124
6.5.2.2 Analysis of student solutions 124
6.5.2.3 Post-test 125
6.5.2.4 Analysis of student solutions 125
6.5.3 Method of direct proof and number theory 126
6.5.3.1 Pre-test 126
6.5.3.2 Analysis of student solutions 126
6.5.3.3 Post-test 127
6.5.3.4 Analysis of student solutions 127
6.5.4 Conclusions based on the results of direct proof 128
6.5.5 Method of ordinary induction (number sequences) 128
6.5.5.1 Pre-test 128
6.5.5.2 Analysis of student solutions 129
6.5.5.3 Post-test ordinary induction (number sequences) 129
6.5.5.4 Analysis of student solutions 130
6.5.5.5 Conclusions based on the results of ordinary induction 131
6.5.6 Method of strong mathematical
induction (recursive sequences) 131
6.5.6.1 Pre-test 131
6.5.6.2 Analysis of student solutions 132
6.5.6.3 Post-test 133
6.5.6.4 Analysis of student solutions 134
6.5.6.5 Conclusions based on the results of strong
mathematical induction 135

iv
  
Chapter 7 Statistical Analysis of Results
 
7.1 Introduction 136
7.2 Logic pre versus Proof pre   137
7.3 Logic post versus Proof post 138
7.4 Logic change versus Proof change   138
7.5 Graphs 139
7.5.1 Logic pre vs Proof pre 139
7.5.2 Logic post vs Proof post 140
7.5.3 Logic change vs Proof change 141
7.6 Control Group versus Experimental Group 142
7.7 Stratified Analysis 144
7.7.1 Controlling for a pre-score of 0 144
7.7.2 Controlling for a pre-score of 1 144
7.7.3 Controlling for a pre-score of 2 145
7.7.4 Controlling for a pre-score of 3 145
7.7.5 Cochran-Mantel-Haenszel statistics 146
7.8 Conclusions based on the comparison between
control and experimental groups 146
7.9 Summary of conclusions on statistical analysis 146

Chapter 8 Suggestions, Recommendations, Conclusion and Future

research

8.1 Introduction 147

8.2 Phases of Proof 147
8.3 Logic and Deductive reasoning 149
8.4 Logic and cues 150
8.5 A Hierarchical and Sequential System for proof 150
8.6 Proof structure and practice 151
8.6.1 Discussion 151
8.7 Assessment and intervention strategies 153
8.8 Mathematical language versus Everyday language 154
8.9 Pedagogical content knowledge 154
8.10 Conclusion 155
8.11 Future Research 155

References 156

v
Appendices  

Appendix A1: Puzzle I (pre-test)   162

Appendix A2: Puzzle III (post-test of puzzle I) 163
 
Appendix B1: Knights and knaves (pre-test) 164
Appendix B2: Knights and knaves (post-test)  165
Appendix C1: Arguments with quantified statements (pre-test) 166
Appendix C2: Arguments with quantified statements (post-test) 167
Appendix D1: Quantified statements - forming conclusions (pre-test) 168
Appendix D2: Quantified statements - forming conclusions (post-test) 169
Appendix E1: Set theory (pre-test) 170
Appendix E2: Set theory (post-test) 171
Appendix F1: Direct proof – divisibility (pre-test) 172
Appendix F2: Direct proof – divisibility (post-test) 173
Appendix G1: Direct proof – number theory (pre-test) 174
Appendix G2: Direct proof – number theory (post-test) 175
Appendix H1: Ordinary induction – number sequences (pre-test) 176
Appendix H2: Ordinary induction – number sequences (post-test) 177
Appendix I1: Strong mathematical induction – recursive sequences (pre-test) 178
Appendix I2: Strong mathematical induction – recursive sequences (post-test) 179
Appendix J: Teachers questionnaire 180
Appendix K: Participant consent form 181
Appendix L: Permission from WCED to do research in schools 182

vi
  

 
Abstract
 
An analysis of South African school mathematics results indicates that one of the
 
problem areas in the mathematical performance of learners is proof and proving. In an
endeavour to improve the mathematical proving ability of first year students at UWC, the
MAM 112 class (a first year elective mathematics course) was taught a course in
elementary logic.

In the initial part of the study, logic puzzles were utilized as a tool to teach students to
make logical connections between and from mathematical statements using the rules of
inference. Subsequently research was done to determine if knowledge and understanding
of logic would translate into improved proving abilities of students.

To put proof and proving into perspective the origins and functions of proof was
explicated and proving in South African schools was investigated. Consequently reasons
are advanced for the dismal high school mathematics results in terms of proof and
possible solutions are discussed.

Recent discoveries of neuroscience are utilized to delineate the brain structures and
cognitive processes involved in learning so as to gain a better understanding of the
learning of mathematics. The findings of neuroscience, cognitive psychology and
educational psychology are employed to elucidate the influence of emotion,
confidence, experience and practice on the learning of mathematics in order to determine
which factors can be applied to improve the proving abilities of students.

The findings of the study indicate that knowledge of logic does help to improve the
ability of students to make logical connections (deductions) between and from
statements. The results of the study, however, do not indicate that knowledge and
understanding of logic translates into improved proving ability of mathematical
statements by students.

vii
  

 
Declaration
 

I declare that
Elementary Logic as a Tool in Proving Mathematical Statements
Is my own work, that it has not been submitted for any degree or examination in any
other university, and that all the sources I have used or quoted have been indicated and
acknowledged by complete references.

Bruce Matthew May November 2008

…………………………..

viii
  

 
Acknowledgements
 
The following people have earned my eternal gratitude for their help, support,
  of this thesis.
encouragement and guidance in the completing

Praise be to my Lord and saviour Jesus Christ the source of my strength and without
whom this thesis would not have been possible.

I will be forever grateful to Prof. R. L. Fray who has presented me with this opportunity
and who with gentle wisdom has guided me patiently.

The Lord sometimes sends people on your path to lighten your load. Prof. R. Madsen is
an example of such a person, his help with the statistical analysis of the data has proven
to be invaluable.

Prof. R. Blignaut has helped to interpret the statistical results. She has done this
unselfishly and without thought of personal gain.

My wife who has sacrificed much has been a pillar of strength and a source of excellent
advice and counsel throughout the research journey.

I will be forever thankful to my mother for her constant prayers, advice, motivation and
unfailing belief in my abilities.

ix
  

 
List of Figures
 

 
Number Description Page

1 Cape Argus article: numeracy and lieracy of grade 3 learners 3

2 Cape Argus article: numeracy and literacy of grade 6 learners 4
3 S.A. mathematics HG paper II (Nov. 2004) 6
4 WCED question analysis mathematics HG paper II( Nov 2004) 7
5 S.A. mathematics HG paper II (nov 2005) 8
6 WCED question analysis mathematics HG paper II( Nov 2005) 9
7 S.A. mathematics HG paper II (nov 2006) 10
8 WCED question analysis mathematics HG paper II( Nov 2006) 11
9 Diagram of valid argument 58
10 Diagram of invalid argument 59
11 Diagram of invalid vs valid argument 60
12 Graph of logic pre vs proof pre 139
13 Graph of logic post vs proof post 140
14 Graph of logic change vs proof change 141

x
  
List of Tables
 
Number Description Page
 
1 Truth table for ~ p   43
2 Truth table for p ∧ q 44
3 Truth table for p ∨ q 44
4 Truth table for ( p ∨ q ) ∧ ~ ( p ∧ q ) 44
5 Truth table for p → q 46
6 Truth table for p ↔ q . 47
7 Truth table valid argument 48
8 Argument analysis of experimental group (pre-test) 97
9 Argument analysis of puzzle I of control group (pre-test) 97
10 Argument analysis of puzzle III of experimental
group (post-test) 100
11 Argument analysis of puzzle III of control group (post-test) 101
12 Comparison of answers of puzzle I and puzzle III 101
13 Argument analysis of knights and knaves pre-test 107
14 Argument analysis of post-test of knights and knaves puzzle 109
15 Comparison of answers to pre- and post-test of knights
and knaves puzzles 109
16 Argument analysis of quantified statements
(question 1) pre-test 112
17 Argument analysis of quantified statements
(question 2) pre-test 112
18 Argument analysis of quantified statements (post-test) 113
19 Argument analysis of quantified statements
(forming conclusions) pre-test 115
20 Argument analysis of quantified statements
(forming conclusions) post-test 117
21 Argument analysis of set theory (pre-test) 121
22 Argument analysis of set theory (post-test) 123
23 Terminology and meanings used in the statistical analysis 136
24 Comparison of Logic pre to Proof pre 137
25 Comparison of Logic post to Proof post 138
26 Comparison of Logic change to Proof change 138
27 Comparison of Pre-test scores of Control and
Experimental groups 142
28 Statistical results based on table 27 142
29 Comparison of Post-test scores of Control and
Experimental groups 143
30 Statistical results based on Table 29 143
31 Controlling for a pre-score of 0 144
32 Controlling for a pre-score of 1 145

xi
33 Controlling for a pre-score of  2 145
34 Controlling for a pre-score of 3 145
35  
Cochran-Mantel-Haenszel Statistics (Modified Ridit Scores) 146
 

xii
 1

 
CHAPTER 1

 
INTRODUCTION

1.1  
The influence of Mathematics on Economic Development
 
The development of good Mathematics and Science teachers and students has been
identified by many as a prerequisite for economic development in countries. The South
African government for example has identified Mathematics, Science and Technology as
areas in education that need investment as a prerequisite for economic growth. The
Dinaledi1 project is an example of such investment by the South African government.
Naledi Pandor (South African minister of education) had the following to say about the
importance of these subjects: “Maths, science and technology are now more important
than they have been in our recorded history.” She went on to say that the importance of
these subjects was also highlighted in the Accelerated and Shared Growth Initiative for
South Africa (Asgi-SA)2 and that Asgi-SA aims to increase economic growth to 6% per
annum between 2010 and 2014, while also halving unemployment and poverty by 2014.
In order to achieve this, critical skills and sectors have been identified. These include the
skills of engineers and the information communication technology (ICT) sector, which
both require a strong knowledge base of mathematics and science. There are however a
number of factors that can and will influence the success or otherwise of such
intervention by government.

1.2 Background to the study

The TIMSS 3 [47] report of 2003 indicates that South African grade 8 learners had the
lowest performance in mathematics and science of the 50 countries that participated in
the study. The study found that in mathematics, South African learners performed
relatively well in the domains of measurement and data, while scoring the lowest in
geometry.
The problems that learners experience in mathematics in South Africa is not restricted to
grade 8, but surfaces as early as in Primary school as can be seen from the following two

1
A crucial initiative arising from the National Strategy for Mathematics, Science and Technology
Education in South Africa was the establishment of the Dinaledi Project in June 2001. As a result of this
project, 102 secondary schools were selected to be centers of excellence for the development of
mathematics, science and technology and was aimed at increasing the participation rates of especially
previously disadvantaged and girl learners and to improve learner performance in these subjects.
2
AsgiSA is a set of government interventions which seek to achieve an average economic growth of 6% by
2010 and to halve poverty and unemployment by 2014.
3
TIMSS (Trends in International Mathematics and Science Study) – is a large-scale comparative study
conducted internationally at the end of the grade 4 and grade 8 year; TIMSS primarily measures learner
achievement in mathematics and science; TIMSS is a project of the International Association for the
Evaluation of International Achievement (IEA); The Human Sciences Research Council has coordinated
and managed the South African part of the study; South Africa was one of 50 countries that participated in
the study; the study was done over four years and 9 000 grade 8 learners from South African schools
participated in the study.
 2

news paper articles4. The first article (shown  in figure 1) is an article that appeared in the
Cape Argus on 22 May 2007 and the second article appeared in the same newspaper on
23 April 2008. The first article reports on a  study that was done by the Western Cape
Education Department (W.C.E.D.). This study was done at 1086 schools and 82 879
learners participated. The study measured the   achievement in numeracy and literacy of
grade 3 learners. The study found that approximately 70% of the grade 3 learners failed
 
to meet the national curriculum requirements for numeracy. The second article reports on
the testing (also done by W.C.E.D.) of 71 847 grade 6 learners in numeracy and literacy.
The learners were tested for ability in numeracy and literacy. The report show that only
14% of grade 6 learners achieved more than 50% in numeracy.

4
The statistics used in the articles is published by the department of education in their Education
Management Information Systems (EMIS) of February 2008
 3

Figure 1
4

Figure 2
 5

Learners are taught proofs in mathematics  without being taught the different elements
that constitute a proof and the different types of proof. The result is that whenever the
 
learners are required to do a proof that they have not rote learned they normally struggle.
This is especially so in the case of Euclidean geometry, (which forms the bulk of the
 
proving experiences of learners) where learners have to substantiate their arguments but
since they do not fully understand the proofs involved, they perform very badly in this
 
section of the work. So it seems that the problems learners had with geometry in grade 8
continue through to the senior grades. This is verified by the fact that in most questions
where learners are required to prove mathematical statements they perform very badly as
shown in the analysis of the grade 12 external exams of 2004, 2005 and 2006 of the
Western Cape Education Department (WCED). The questions and question analysis of
2004 is shown in figure 3 and 4; the questions and question analysis of 2005 is presented
in figure 5 and 6; the questions and question analysis of 2006 is shown in figure 7 and 8.
 6

Figure 3
SOURCE: Mathematics HG paper 2 – Nov 2004 (Department of Education RSA)
 7

SOURCE: WCED Examination Results 2004

Figure 4
 8

Figure 5
SOURCE: Mathematics HG paper 2 – Nov 2005 (Department of Education RSA)
 9

SOURCE: WCED Examination Results 2005

Figure 6
 10

Figure 7
SOURCE: Mathematics HG paper 2 – Nov 2006 (Department of Education RSA)
 11

SOURCE: WCED Examination Results 2006

Figure 8

It has been my experience that high school mathematics teachers blame the primary
school teachers for the bad mathematics results. They argue that the primary school
teachers allow the learners to progress despite the fact that the learners did not exhibit the
desired mathematical competencies. Similarly university lecturers blame high school
teachers for bad mathematics results at first year level. I think it is significant that each
succeeding level of the educational system blames the previous level, since it alludes to
the fact that students pass through the system with less than the expected level of
mathematical knowledge and skills. This is especially so in the case of proof and proving
as can be seen from the question analysis.

Kutzler [58] is of similar opinion as can be seen from the following quote: “In
mathematics teaching at school we simply don’t have enough time to wait until all
students have completed all previous storeys. The curriculum forces the teacher to
continue with the next topic, independent of the progress of individual students.” The
grade 12 mathematics results of the WCED (fig 4, 6 and 8) bears testimony to the fact
that in South Africa this is also the case i.e. that some learners get passed on from grade
to grade without acquiring the requisite skills and knowledge for the next grade. For
example if a learner does not develop the competency to solve linear equations, then such
a learner will struggle with the solving of quadratic equations, since steps in solving
quadratic equations requires the solving of linear equations. This in turn will cause the
learner to struggle with finding the x-intercepts of graphs of parabolas as this is
dependent on solving quadratic equations, etc.

Add to the above argument the fact that in the majority of cases in mathematics
assessment is utilized only for progression purposes and very rarely as a diagnostic tool
(i.e. to determine and solve reasoning problems of students). The result is that students
pass right through the educational system without their mathematical reasoning problems
being addressed. I, for example, have discovered that some grade 12 learners still struggle
 12

with addition and subtraction of fractions.  This is an example of the difficulty that
learners develop with some aspects of numeracy at primary school level (as indicated in
 
figures 1 and 2) and which is often also exhibited at high school level. This then alludes
to the fact that numeracy problems such as addition and subtraction of fractions might
have been uncovered by assessment at primary   school, but has never been addressed.

 
The TIMSS [47] report of 2003 indicated that South African mathematics and science
teachers are among the least qualified of the 50 countries that participated in the study.
This can be a contributing factor of the low marks obtained by learners in questions on
geometry in the WCED Examinations given in figures 4, 6 and 8.

1.3 The Aim of the study

The International Commission on Mathematical Instruction (ICMI) 5 [48] will have a

study conference that will deal with the role of proof and proving in mathematical
education in 2009 in Taipei in Taiwan. In their discussion document for this study, the
organizers of the conference stated the following in connection with the role of logic in
proof and proving. “The traditional assumption has always been that teaching students
formal logic would easily translate into helping them understand the deductive structure
of mathematics, and also to write proofs. But as research has shown this transfer doesn’t
happen automatically. It is still not clear what benefit, if any, may arise from teaching
formal logic to students or to prospective teachers, in particular when mathematicians
have been known to readily admit that they infrequently use formal logic in their
research. Hence there is a need for more research that might provide support for the
conclusion that the teaching of formal logic increases students’ ability to prove or to
understand proofs.”

As a result of the above arguments and in an attempt to improve the ability of first year
university mathematics students to read, understand and prove mathematical statements,
we have embarked on teaching these students a course in mathematical logic in the hope
that this will improve their mathematical proving abilities.
What we are not claiming is that we will get all the students to have the same ability i.e.
that we will get all of them to become successful in proving mathematical statements.
This is because not all students develop the same abilities while learning. Vygotsky puts
it as follows: “When it was first shown that the capability of children with equal levels of
mental development to learn under a teacher’s guidance varied to a high degree, it
became apparent that those children were not mentally the same age and that the
subsequent course of their learning would obviously be different.” We are also not
claiming that the reasons advanced by us are the only reasons responsible for the bad
mathematical results at school, we accept that there might be other factors involved that
we have not addressed.

5
The International Commission on Mathematical Instruction (ICMI) was established in 1908. The
members of ICMI are neither individuals nor organizations, but countries. The ICMI currently consists of
72 member states. The focus of the ICMI is to enquire into mathematics teaching in countries world-wide.
 13

1.4 The Main Research Question  

  answer is:
The main research question this study wants to
  taught a course in elementary logic,
If first year university mathematics students are
would the acquired competency translate into increased ability to prove mathematical
 
statements?

In order to answer this question we taught first year mathematics students at the
University of the Western Cape a course in elementary logic. To determine their initial
reasoning abilities we gave the students logical puzzles to solve without any prior
teaching. We considered this as a pre-test. The students then received instruction on
elementary logic after which they were given similar puzzles to solve which was
considered as a post-test. This was done to ascertain if the students have acquired the
ability to apply their knowledge. Subsequently the students received instruction on
different types of proof and on conclusion of each type of proof were presented with a
post-test. These pre- and post-tests were then analyzed statistically to establish if there
was an improvement in proving ability.
 14

 
CHAPTER 2

  REVIEW
LITERATURE

2.1 Introduction  

 
The literature shows that a number of studies were done to determine the effects of a
course in logic on student proving abilities. Some of these studies were done at high
school whilst others were done in the first year at university level. The studies done with
high school learners were done in the domain of Euclidean geometry. Four articles that
investigated the effect of logic on proving at high school will be reviewed, whilst two
articles will be reviewed that investigated the effect of logic on proving at tertiary level.

2.2 Effect of logic on proving at school level

The purpose of the study by G.W. Deer [22] of Florida State University was to determine
what effects the teaching of an explicit unit in logic would have on students’ ability to
write proofs in high school geometry and in an unrelated mathematical system in which
the students had no previous training. To accomplish this Deer divided a high school
geometry class into two groups of 13 students each. One group was taught a unit in
introductory logic including the basic inference patterns. The other group studied material
on games, puzzles, codes and modular arithmetic during the time the first group was
taught logic. At the time of the study the students had not been introduced to the concept
of proof in geometry. Pre- and post- tests were administered to measure the students’
understanding of the basic inference patterns in logic. The unit in logic, including the two
tests, involved nine days of class work.

After these nine days the two groups were combined into one class and taught to prove
theorems in geometry. Lessons on congruent angles, supplementary and complimentary
angles, perpendicular lines, right angles and the intersection of lines was used to present
the proof concept. The types of proof used were direct and indirect proof and the two-
column format was used for all proofs. The axioms necessary for these lessons were
introduced prior to the study. The unit in proof writing was done over nine days of class
work.

Two tests were given upon completion of the unit in proof writing. One was a test of their
ability to write proofs of the theorems in geometry. This test consisted of theorems the
learners had not seen before, but were easily derived from the material studied. The other
was a test of their ability to write proofs in a mathematical system in which they had no
prior experience. For this test a set of 7 axioms was given and then using these axioms
the students were asked to prove 5 theorems. Each test was done over an hour.

This study found that there was no significant difference in the performance of the two
groups on each test. Hence the conclusion of the study was that a course in logic does not
enhance the ability to write proofs in high school geometry and therefore logic does not
improve the ability to do deductive reasoning.
 15

The study of John Lewis Platt [70] of Colorado   state college was designed to evaluate the
effect of the use of mathematical logic in high school geometry on: achievement of
students in geometry; achievement of students   in reasoning in geometry; critical thinking
of students and attitude of students toward logic deduction and proof in mathematics.
Two groups of six classes were involved in  the study. One group was an experimental
group of six classes who participated in a four week unit in mathematical logic and who
 
applied the logic in the learning of deductive methods in geometry. The control group
was taught the same course in geometry but without the study of mathematical logic. The
conclusions of the study were as follows: mathematical logic is an appropriate area of
study for both high and average achievers in high school geometry; there is no loss of
achievement in geometry caused by devoting time from the traditional course in
geometry to the teaching of mathematical logic; including mathematical logic in high
school geometry does not result in a course which is significantly superior to the
traditional course in its over-all effect upon student achievement in reasoning in
geometry, critical thinking ability or attitude of students toward logic, deductive thinking
and proofs in mathematics; including instruction in mathematical logic appears to
produce a more effective treatment of high school geometry with high achieving students
in its effect upon student achievements in reasoning in geometry.

The study conducted by David Mueller [67] was intended to determine the effects of the
teaching of a logic unit on the proof writing abilities of high school geometry students in
familiar and unfamiliar settings. The study was done at a Florida high school in the first
half of the 1973 – 74 school year and involved four teachers, six classes and 146 students.
Two classes were reserved for the higher ability students and the remainder of the classes
had a regular mixture of ability levels. The two classes were treated differently. The
higher ability group first studied the logic unit and then two geometry units, whilst the
other group studied the logic unit between the two geometry units. The two geometry
units consisted of chapters 1, 4 and 5 from the book on geometry by Moise and Downs
(1964) which was the prescribed book. The logic unit was taken from a book written by
Bastic in 1969. This book was unusual in three main ways, namely: it presented the logic
informally rather than in formal syllogisms; it contained material in the form of Lewis
Carrol-type puzzles to assist in learning to draw conclusions and to determine which step
comes next, which are important items in writing proofs; it contained material to
specifically assist in using the interpreting axioms; it made liberal use of nonsense
content in the examples.

To verify that the students learned the logic material a test was given before and after the
logic unit. A test of ability to interpret and use axioms was given before, during and after
the logic unit to provide a check on the effectiveness of materials in the unit designed to
increase this ability. An attitude scale was utilized to evaluate changes in attitude towards
mathematics. Tests of proof writing ability in both familiar (using geometry content just
studied) and unfamiliar settings were administered after each of the two geometry units in
order to test the effects of the logic unit on proof-writing ability.

The main research questions were concerned about the effect on proof-writing skills. The
finding was that the logic unit was of little significant help in writing proofs of theorems
 16

in either the familiar or unfamiliar settings.  Another finding was that the logic unit
appeared to have some initial negative effect, but some beneficial effect later on. The
 
results however were mixed for the various classes and showed no completely consistent
pattern. The minor research questions concerned the change in attitudes and ability to
interpret and use axioms. One of the findings  was that there was some decline in attitude
toward mathematics, but it could not be determined whether it was due to the logic unit
 
or at least partially to other causes. The other finding was that the last part of the logic
unit very significantly aided in improvement of the ability to interpret and use axioms as
it was intended to .

The general conclusion was that the logic unit was only partially successful in answering
the research questions. The researcher therefore recommended that a more effective
approach would be an integrated approach. He suggested that this should include
presentation of the key logic techniques and the use of explicit reminders that these
techniques are embodied in the examples.

Epp [34] in her summary of the Texas Pre-freshman Engineering program (Tex Prep) of
Berriozabal [5] states that the study reports considerable long-term impact of the study of
logic on the students in the program. This was a comprehensive program intended to
identify middle and high school students with the potential and interest in becoming
engineers and scientists and to guide them toward acquiring knowledge and skills
required for success in their professional aspirations. Prep was an academically intense
eight week summer program which stresses the development of abstract reasoning skills,
problem solving skills and their application. The report indicates that since 1979, 8067
students have successfully completed at least one summer of Tex Prep. The program was
run over 3 years and in the first year logic and its application to mathematics was done.
The research was done by means of questionnaires.

2.3 Effect of logic on proving at tertiary level

The aim of the study done by R.L. Walter [90] of Florida state university was to
determine the effect of knowledge of logic in proving mathematical theorems in the
context of mathematical induction. His study was aimed at determining if knowledge of
logic would cause people to perform better on tests on the principle of mathematical
induction than people that have no knowledge of logic. In order to determine this,
instructional material in logic and the principle of mathematical induction were presented
to the students.

The logic unit was not programmed, whereas both of the mathematical induction units
traditional and experimental were programmed. The experiment was run twice, using pre-
calculus college students first and college calculus students the second time. The test that
was administered consisted of two parts, namely, multiple choice and proof. In both cases
the results were in favour of the research hypothesis, however only the variable proof
with pre-calculus students was significant at the .05 probability level. The results of the
calculus students on the other hand were far from significant with probability values
greater than .50.
 17

Using the ACT6 English, mathematics and  composite scores to define ability levels, it
was apparent that the students from the calculus classes had a higher ability level than
those students in the pre-calculus classes. In  the pre-calculus class students had a wider
range of abilities than those from the calculus classes. The investigator therefore
 
concluded that the effect due to different treatments was not as evident in the calculus
students because of the high ability level present. In other words the calculus students did
 
well in spite of the treatment. In view of the results obtained with both the pre-calculus
and the calculus students the researcher concluded that the research hypothesis was not
without merit since in both cases the results favoured the research hypothesis. He
therefore suggested further research into the effect of the knowledge of logic on different
instructional strategies in the principle of mathematical induction.

Cheng et al [14] are of the opinion that two views have dominated theories of deductive
thinking. One is the view that people reason using syntactic, domain independent rules of
logic and the other is the view that people use domain-specific knowledge. In contrast to
the above two views Cheng et al are of the opinion that people often reason using
pragmatic reasoning schemas. These schemas are clusters of rules that are highly
generalized and abstracted but nonetheless defined with respect to classes of goals and
types of relationships. The researchers therefore designed three experiments, using
college students, to examine the processes involved in deductive reasoning.

In experiment 1 it was investigated how people interpret and reason about a type of
logical statement, called the conditional. The researchers randomly assigned 80 students
(in equal groups) of the University of Michigan to each of four training groups namely:
rule training; examples training; rule plus examples training and; no training. None of the
chosen students had previously received any formal training in logic.

The training materials and methods were as follows: for rule training the students
received a seven-page booklet containing an exposition on conditional statements
followed by an inference exercise. The exposition consisted of an explanation of the
equivalence between a conditional statement and its contra-positive, as well as an
explanation of the two common fallacies of affirming the consequent and denying the
antecedent. The contra-positive was explained in part by the use of a truth table and in
part by Euler diagrams that used concentric circles to show the relations between a
conditional statement and its contra-positive and in part by an illustrative conditional
statement. Students were given immediate feedback on correctness of the inference
exercises as well as a brief explanation of the correct answer.

With the examples training students were requested to solve two selection problems7.
Students were given immediate feedback about their performance. The Rule plus
Example training consisted of the materials for the rule condition followed by those for

6
American College Testing (ACT) – Is a national USA college admission and placement examination. It
consists of four tests namely English, Mathematics, Reading and Science. The score range for each of the
four tests is 1 – 36. The composite score as reported by ACT is the average of the four test scores earned
during a single test administration rounded to the nearest whole number.
7
Wason’s selection task was used by the researchers.
 18

the examples condition. The only further   addition was that for these students the
explanation of the correct answer for each example was given in terms of the abstract
rules they had just learned. The students were   also given eight selection problems as a
test. This test consisted of two of each of three types of problems involving a
bi-conditional rule.  

 
The finding of experiment 1 was that training in standard logic when coupled with
training on examples of selection problems leads to improved performance on subsequent
selection problems. In contrast, training on rules of logic without such examples failed to
significantly improve performance. The researcher therefore concluded that this is
consistent with their view that the material conditional is not part of people’s intuitive
reasoning abilities and that because of this they lack any ability to put abstract rule
training to use.

With experiment 2 the researchers examined the impact of a one-semester undergraduate

course in standard logic. The students of two introductory logic classes were used as
subjects. One class was held at the Ann Arbor campus of the University of Michigan and
one at the branch campus at Dearborn. Both classes received 40 hours of lectures and
covered topics in propositional logic, including modus ponens, modus tollens, affirming
the consequent and denying the antecedent, and the distinction between the conditional
and the bi-conditional. The textbook used in one class was Elementary logic by Simco
and James and in the other it was Introduction to Logic by Copi. The course utilized
meaningful conditional sentences to illustrate the inference rules and fallacies. The
emphasis however was placed on formal logical analyses i.e. truth-table analyses and
construction of proofs.

A pre-test and post-test were administered to each class. The pre-test was given in the
first week of class before any discussion had taken place. The post-test was given in the
final week of the semester. The eight selection problems used in experiment 1 were
divided into two sets of four and were given in a booklet to students to complete during
lectures. No feedback was provided until after completion of the post-test. Only data
from students who completed both the pre-test and the post-test were analyzed, so that
the effect of logic training could be treated as a within-subject variable. A total of 53
students completed the study.

The findings was that abstract training in the logic of conditional statements does not
have much effect on the way people reason about problems that could potentially be
solved by its use. On the other hand problems that can be interpreted in terms of
pragmatic reasoning schemas were solved by a large proportion of the students. The
researchers therefore were of the opinion that it might be possible to improve people’s
deductive reasoning by training them on pragmatic reasoning schemas.

In experiment 3 the researchers wanted to test the above possibility. For this experiment
72 University of Michigan undergraduates were selected. The students were randomly
assigned in equal numbers, to one of three groups. One group was assigned a control
condition in which no training was given, whereas the second group was assigned to an
 19

obligation-training condition and the third to  a contingency-training condition. Students

that were assigned to the two training conditions were given the appropriate training
materials to read for 10 minutes and were then  asked to complete the test problems. The
control subjects, i.e. the students that received no training, were simply given the test
problems and asked to complete them.  

 
The obligation training material consisted of a two-page booklet that contained details on
the nature of obligations, the procedures necessary for checking if a violation of the
obligation has occurred and also an example of an obligation statement presented for
assessing obligations were described in terms of four rules, one for each of the four
possible situations that might arise, and were represented by p, not p, q and not q.

The contingency training material was similar to the obligation-training material except
that the checking procedures were described in terms of assessment of “contingencies”
rather than “obligations”. The test- problem booklet consisted of eight selection
problems. Four of these problems were easily discernible as obligation situations and the
other four were arbitrary. The finding of this experiment was that it was effective for at
least the arbitrary problems.

Based on the above three experiments the researchers concluded that people typically
reason using abstract knowledge structures organized pragmatically, rather than in terms
of purely syntactic rules of the sort that comprise standard logic. They found that students
reasoned in closer accord with standard logic when thinking about problems intended to
evoke permission or obligation schemas than when thinking about purely arbitrary
relations.

The training results showed that an entire course in standard logic had no effect on the
avoidance of any error except a slight reduction in the fallacy of affirming the
consequent. A brief training session in formal logic of a type that was proved to produce
substantial effects on people’s ability to reason using the law of large numbers, had no
significant effect on students’ ability to use modus ponens or modus tollens or to avoid
the errors of affirming the consequent or denying the antecedent. The researchers
however was of the opinion that rule training was not useless, since when it was
combined with examples training, students were able to make substantial use of it. They
furthermore concluded, based on results that since modus ponens are not a universal rule
of natural logic that therefore it is highly unlikely that any formal deductive rule is
general across the adult population.

The researchers stress that one of the educational implications of the study is that
deductive reasoning is not likely to be improved by training on standard logic. They
however also state the following: “ …if logic instructors wish to influence their students’
inferential behaviour in the face of novel problems, they must do much more than they
currently do to show how to apply logical rules to concrete problems.” So it seems that
by providing guidance as to where and how logic should be applied might subvert this
educational implication.
 20

2.4 Literature review summary  

 
All three articles (Deer [22], Platt [70], Mueller [67]) dealing with high school geometry
concluded that logic does not help to improve proof-writing of high school students. Platt
however also conclude that logic does help   teachers to deal more effectively with
geometry and that it is an appropriate area of study for both high and average achievers.
 
Mueller found that logic did help students to improve in their ability to interpret and use
axioms in proving. He recommended an integrated approach to the teaching of logic, that
include the presentation of logic techniques and the use of explicit reminders as to where
and when to employ logic.

Cheng et al [14] concluded in their study with university students that training on logic
does not improve deductive reasoning and hence proving abilities of students. They do
however show that in some instances training on pragmatic reasoning schemes helped to
improve the deductive reasoning of people. Another suggestion is that instructors at
tertiary level should show their students how to apply logical rules to concrete problems
in order to enhance their inferential behaviour.

On the contrary the study done by Walter [90] with college students showed that logic
does help students to improve their proving abilities. The study however only researched
the effect of logic on proof in the context of mathematical induction. This study therefore
paved the way for research into the effects of knowledge of logic on proving abilities
using other methods of proof used at tertiary institutions. Hence our study researched the
effect of knowledge of logic on direct proof, contradiction, contraposition and
mathematical induction in the context of number theory and set theory.

The study by Mueller [67] made use of Lewis-Carol-type logic puzzles to assist the
students to learn to draw conclusions and to determine which step comes next in writing
proofs. Our study also made liberal use of this informal way of introducing proof
techniques to students and one of the aims of the study was to determine if this method
was effective in helping students to improve their deductive abilities.
 21

 
CHAPTER 3

THEORETICAL  FRAMEWORK

3.1 Introduction  

 
To put proof and proving in perspective in the current educational and research
environment, I am of the opinion that one should start with some elementary questions
like: What is proof in mathematics? Is proof important in mathematics and if so why?

3.2 Is proof important in mathematics and if so why?

Let us start with the second question that refers to the importance of proof in mathematics
since it seems to be less contentious. There seems to be general consensus that proof
plays a central role in mathematics. Jones and Rodd [50] of the British Society for
Research into Learning Mathematics state: “Proof and proving are of course, central to all
mathematics”. Volmink [88] advances the following argument as to the importance of
proof: “It may be asserted that proof is the essence of mathematics”. Knuth [53] is also
convinced that proof is very important in mathematics as the following quote shows:
“Many consider proof to be central to the discipline of mathematics and the practice of
mathematicians.” Herbst [44] answers the question why proof is important in
mathematics education as follows: “Proof is essential in mathematics education not only
as a valuable process for students to engage in (such as developing their capacity for
mathematical reasoning), but more importantly as a necessary aspect of knowledge
construction.” Stylianides and Stylianides [80] advances the following reasons as to the
importance of proof: “…proof is fundamental to doing mathematics – it is the basis of
mathematical understanding and is essential for developing, establishing, and
communicating mathematical knowledge. Secondly, students’ proficiency in proof can
improve their mathematical proficiency more broadly because proof is involved in all
situations where conclusions are to be reached and decisions to be made.” Also proof is
seen by some to go hand–in-hand with the deductive reasoning ability of students as can
be seen from the following quote in Stylianides and Stylianides: “In addition, from a
psychological standpoint, the development of students’ ability for deductive reasoning
has been found to go along with the development of their ability for proof.”

The question as to what proof is, is not as straightforward as there seems to be conflicting
views as to what proof is. The following quote from Kleiner [52] based on Gödel’s
incompleteness theorems make this abundantly clear: “Gödel’s work demands the
surprising and for many discomforting conclusion that there can be no definitive answer
to the question “what is proof?”. Reid [71] is convinced that there is no consensus among
mathematics researchers and teachers as to what proof is and he therefore states: “But if
we can acknowledge that there is a problem and discuss the characteristics of proof, we
may be able to come to, if not agreement, then at least agreement on how we differ.”
 22

3.3 Origins of proof  

  proof, I believe one needs to start at the

In order to explain and understand the concept
origins of proof. There is general consensus among the mathematical history pundits that
the first people that introduced the concept   of mathematical proof were the ancient
Greeks (Kleiner [52], Kutzler [58], Barker [3], Stillwell [79]). There are however other
 
nations that contributed enormously to the mathematical knowledge, concerning proof
that is available today. Two of these nations are the Egyptians and Babylonians. Their
mathematics however lacked the concept of proof (Kleiner [52] ; Barker [3]; Kutzler[58])
and was based on inductive reasoning. But what is inductive reasoning?

According to Ayalon and Even [2] there are various sorts of thinking and reasoning
including creation, induction, plausible inference and deduction. For our purposes we are
going to elucidate only inductive and deductive reasoning. Ayalon and Even defines
inductive reasoning as follows: “…developing hypotheses based on empirical
observations to describe “truths” or “facts” about our world.” They define deductive
reasoning as follows: “ Deductive reasoning is unique in that it is the process of inferring
conclusions from known information (called premises) based on formal logic rules,
where conclusions are necessarily derived from the given information and there is no
need to validate them by experiments.” Stylianides and Stylianides [80] concur with this
definition of deductive reasoning. They see it as the general form of reasoning associated
with logically necessary inferences based on given sets of premises.

One way of putting the above methods of reasoning into perspective is to look at
Buchberger’s Creativity spiral as explained by Kutzler [58]. Buchberger’s spiral provides
the different aspects that are involved in the discovery of mathematical knowledge. In the
spiral there are three phases namely experimentation, exactification and application.
Kutzler explains that during the phase of experimentation one applies known algorithms
to generate examples then obtains conjectures through observation. During the phase of
exactification conjectures are turned into theorems through the method of proving, then
algorithmically useful knowledge is implemented as algorithms. During the phase of
application one applies algorithms to real or fictitious data. Using the above explanation
Kutzler asserts that the Egyptians and other ancient civilizations applied only the phases
of experimentation and application in their construction of mathematical knowledge. So
they only used inductive reasoning. He argues that in about 500 B.C. the Greeks took the
Egyptian mathematics and applied to it the deductive method of reasoning i.e. they added
the phase of exactification. He maintains that from then on mathematics comprised of all
three phases and that mathematics was thus established as the deductive science of today.

However from about 1950 on the French mathematician Dieudonne and his colleagues
(known as the Bourbaki group) developed the system of “definition-theorem-proof-
corollary-…” The Bourbaki system therefore did not include the phase of
experimentation and consists only of the phases of exactification and application. This
Bourbaki system has now become part of the modern process of teaching and learning.
Thus it has become customary to teach mathematics by deductively presenting
 23

 
mathematical knowledge and then asking students to learn it and then use it to solve
homework and examination problems.
 
Since the deductive method is prevalent in mathematics, I think it is imperative that it is
  to start would be with geometry, since the
subjected to closer scrutiny. The obvious place
Greeks that started the deductive method is synonymous with geometry. There were a
 
number of Greek thinkers that gave attention to geometry, but it was Euclid that
presented in systematic form, all the main geometrical discoveries of his predecessors in
his classic book “The Elements”. This book played an influential role in Western thought
as can be seen from the following quote from Barker [3]: “ Through ancient times,
through the medieval era, and in the modern period right up into the nineteenth century,
Euclid’s Elements served not only as the textbook of geometry but also as a model of
what scientific thinking should be.” I am of the opinion that an important question now
is: What are the salient features of Euclid’s procedures that distinguishes it from the
inductive method?

The nations that applied the inductive method were satisfied with showing that a
principle holds for a number of particular cases by means of experimentation and
observation. This however did not show that it is true in general but only for the
particular cases studied. Conversely a distinctive feature of Euclid’s Elements is that he
formulates his geometrical laws in universal form i.e. he proves that his laws are true in
ALL cases. Barker [3] states of Euclid in this regard: “He never concerns himself with
actual experiments or observations like that. Instead his proofs are deductive proofs by
means of which he seeks to establish his conclusions with the rigor of absolute logical
necessity.” But what is this proof that was now being introduced as a mathematical
procedure?

Barker [3] defines proof in the classic sense as follows: “A proof is a chain of reasoning
that succeeds in establishing a conclusion by showing that it follows logically from
premises that already are known to be true. We cannot have a proof unless we can start
with one or more already known premises which serve as a basis upon which the proof is
to rest.” What Euclid did in order to use known premises to start his proofs from was to
divide the geometrical laws into two groups. The postulates are a small group of laws
which Euclid thought are self-evident and therefore need no proving, but which he
adopted as basic premises. Examples of these are laws about lines, angles and figures,
laws which Euclid regarded as true and which he employed for the proof of other laws.
Conversely Euclid assumed that there were an infinitely large group of other geometrical
laws which can be proved using these postulates. This second group of laws he named
theorems or propositions.

Euclid did not only deal with geometrical laws, but also with principles that deal with
equality of magnitude. These five other principles were called axioms. A well known
example of one of these axioms is: “the whole is greater than the part.” In the modern era
the distinction between axioms and postulates has become blurred and today the words
axioms and postulates are used interchangeably. All of the above discussion about
Euclid’s methodology is succinctly put by Barker as follows: “The postulates, axioms
 24

and definitions supply the starting point for   Euclid’s proofs. His aim is to prove all his
other geometrical principles, first those of plane geometry and then later on those of solid
 
geometry, by showing that they follow necessarily from the basic assumptions”. So in the
Euclidean system the axioms were regarded as self-evident truths and therefore there was
 
no need to prove them. However with the development of the non-Euclidean geometries
of Lobachevsky, Bolayi and Riemann it became clear that not all propositions within a
 
formal system can be proved without getting circularity and that consequently certain
propositions had to be accepted as starting points. De Villiers [29] therefore argues as
follows: “Whereas many had previously believed that axioms were “self-evident truths”,
they now realized that they were simply “necessary starting points” for mathematical
systems.”

The above then is how proof in mathematics was started. Since then however the concept
of proof has been vigorously debated. In the 19th century there were frequent
disagreements among mathematicians concerning what they thought what the
foundations of mathematics was. In the early 20th century this disaffection among
mathematicians gave rise to three schools of mathematical thought. These schools of
thought gave rise to the philosophies of logicism, formalism and intuitionism.
Kleiner [52] and Hanna [42] are convinced that this was the first formal expression by
mathematicians of what mathematics is about and in particular, of what proof in
mathematics is about.

The logicists advocated that mathematics is part of logic. To them mathematical concepts
are expressible in terms of logical concepts and mathematical theorems (proofs) are
tautologies i.e. are true by virtue of their form rather than content
(Kleiner [52], Hanna [42])

The formalists viewed mathematics as a study of axiomatic systems. According to this

philosophy both the primitive (undefined) terms and the axioms are considered to be
strings of symbols to which no meaning is to be attached. These primitive terms and
axioms are to be manipulated according to established rules of inference to obtain the
theorems of the system. (Kleiner [52])

The intuitionists on the other hand claimed that no formal analysis of axiomatic systems
is necessary. They believed that the mathematician’s intuition, beginning with that of
number, should guide him to avoid contradictions and that he must pay special attention
to definitions and methods of proof. For the intuitionists mathematics and mathematical
language are two separate entities with mathematical activity essentially a languageless
activity of the mind. They argue that their methods of proof must be constructive and
finite in nature. (Kleiner; Hanna)

3.4 Functions of proof

According to Griffiths [41] the traditional view of proof is as follows: “A mathematical

proof is a formal and logical line of reasoning that begins with a set of axioms and moves
through logical steps to a conclusion.” In this view of proof there is an emphasis on form
 25

and systematization i.e. starting with axioms  and then by reasoning deductively coming
to a conclusion. This view however does not enjoy complete support among
mathematicians and mathematics educators today.  This lack of support in some quarters
is confirmed by Hanna [43] in the following quote: “In the last two decades several
mathematicians and mathematics educators  have challenged the tenet that the most
significant aspect of mathematics is reasoning by deduction culminating in formal
 
proofs…Mathematicians agree, furthermore that when a proof is valid by virtue of its
form only, without regard to its content, it is likely to add very little to an understanding
of its subject and ironically may not even be very convincing.” As a result of this
disagreement with the traditional view of proof, a more recent view of proof is that proof
is an argument needed to validate a statement, an argument that may assume several
different forms as long as it is convincing (Hanna [43] ). This alternative view of proof,
that does not only depend on logical structure, but also on how convincing it is, enjoys
different interpretations as to who must be convinced. Weber [91] supplies some of these
alternative views. He argues that some of the mathematical fraternity (Maso, Burton and
Stacey) see proof as an argument that must convince an enemy. Others such as Davis and
Hersch see proof as an argument that must convince a mathematician who knows the
subject, while Volmink [88] and Epp [34] sees proof as an argument that must convince a
reasonable skeptic. Balacheff and Manin focused on the social and contextual nature of
proof. Balacheff define proof as an explanation accepted by a given community at a
given time, while Manin argues that an argument becomes a proof only after the social
act of accepting the argument as a proof.

It would seem, based on the above arguments about proof, the only function of proof is
that of convincing. There are some however that assign functions other than convincing
to proof. The following list of functions of proof is provided by Hanna [43]: verification
i.e. concerned with the truth of a statement; explanation i.e. providing insight into why it
is true; systematization i.e. the organization of various results into a deductive system of
axioms, major concepts and theorems; discovery i.e. the discovery of new results;
communication i.e. the transmission of mathematical knowledge; construction of an
empirical theory; exploration of the meaning of a definition or the consequences of an
assumption; incorporation of a well-known fact into a new framework and thus viewing it
from a fresh perspective. The development of mathematical intuition is seen as another
function of proof by Pinto and Tall. Yackel and Cobb are convinced that by teaching
students how to prove they will develop the ability to independently construct and
validate new mathematical knowledge. They are therefore of the opinion that teaching
students how to prove will cause them to become autonomous in mathematics.

Reid [71] argues that there is a dimension of proof that is different from the dimensions
of concept and purpose that we need to consider. He identifies this as the reasoning
involved in proving. He argues that there are three prevalent views on the kind of
reasoning involved in proving. These are that either there are several distinct kinds of
reasoning involved in proving or several kinds of reasoning in combination are involved
in proving or deductive reasoning alone constitutes proving. He argues further that those
with a traditional concept of proof would be more likely to see proof as purely deductive,
 26

while those with a quasi-empirical concept  of proof might be more inclined to view
proving as a combination of kinds of reasoning.
 
3.5 Proving in South African schools
 
Since our study targeted first year students it is imperative that we determine what their
 
previous experience with proof and proving has been at school level.

In South Africa a new curriculum known as the National Curriculum Statement (NCS)
was implemented in grade 10 in 2006. Our target group of students however still dealt
with the old curriculum so we are going to look at proof and proving in the old
curriculum. In this curriculum Euclidean geometry usually was the topic used to
introduce formal proof to learners. In South Africa the two column-proof format is used
extensively in Euclidean geometry. But what is a two-column proof and why is it used?
Herbst [45] argues that Arthur Schultze and Frank Sevenoak were the first to use the two-
column format for writing out proofs in geometry. In this format statement and reasons
were written next to each other divided by a vertical line. Herbst quotes Schultze and
Sevenoak to elucidate the use of the two-column proof: “Every proof consists of a
number of statements each of which is supported by a definite reason. The only
admissible reasons are: a previously provided proposition; an axiom; a definition; or the
hypothesis.” The justification provided for giving reasons for each statement is that the
reasons given by the student will enlighten the teacher as to the understanding of the
student of the deductive argument involved in the composition of the proof.

The geometry courses of the USA were composed of two different kinds of propositions
namely Fundamental propositions and exercises. The Fundamental propositions are
described as the minimum which all pupils should know. These Fundamentals would
serve to develop the subject of studies and to exemplify what it meant to prove a
proposition. On the other hand the Fundamentals themselves would be the information
that students would use as they did proof exercises. Herbst [45] maintains that these
exercises were meant to stimulate student reasoning and to practice what had already
been learned. South Africa has a similar system. In South Africa the Fundamental
propositions are the prescribed theorems and axioms. These prescribed theorems and
axioms formed the underlying knowledge that each learner has to know in order to do
proof exercises. It was expected of the teacher to present proofs of these theorems in the
statement and reason format to learners. Some of the reasons advanced for this format,
was so that students could understand the deductive reasoning involved and also to serve
as an example of how proving should be done. De Villiers [28] writes about this as
follows: “The rationale for including formal geometry in the school curriculum is two-
fold: it is seen as a vehicle for teaching and learning deductive thinking “proof” and also
as a first encounter with a formal axiomatic system.” It is therefore not surprising then
that some teachers and learners at school are under the impression that this is the only
kind of proof there is, and that all proofs should be done in this way as the two-column
proof form the bulk of their proving experiences at school level. In fact research done by
us with 50 teachers in one of the EMDC’s has shown that only two out of the 50 teachers
 27

knew of other types of proof.8 Weber [91]  writes the following in this regard: “One
reason that university students find proof so difficult is that their experience with
  school geometry.” Stylianides et al [80] is
constructing proofs is typically limited to high
of similar opinion as can be seen from the following quote: “… several researchers have
 
identified students’ abrupt introduction to proof in high school as a possible explanation
for the many difficulties that secondary school and university students face with proof,
 
thereby proposing that students engage with proof in a coherent and systematic way
throughout their schooling.”

In 1996 De Villiers [27] claimed that it is a well known fact that learners in grade 12 in
South Africa perform much worse in Euclidean geometry than in algebra. Eleven years
later in 2007 it was still the case as we showed in the introduction in the tables of grade
12 results. Problems with proving are not restricted to grade 12, but become apparent in
grade 8 and continue right through the system. Internationally Euclidean geometry has
also been a thorn in the side of learners. For example De Villiers [27] writes the
following of the Russians: “ In the late sixties Russian researchers undertook a
comprehensive analysis of both the intuitive and systematization phases in order to try
and find an answer to the disturbing question of why pupils who were making good
progress in other school subjects, showed little progress in geometry.” The question is
why do learners struggle so much with Euclidean geometry?

The Van Hiele theory provides some answers to the above question. Pierre van Hiele and
Dina van Hiele-Geldof were a husband-and-wife team of Dutch researchers who noticed
that their students had difficulties in learning geometry. Their observations led them to
develop a theory involving levels of thinking in geometry that students pass through as
they progress from merely recognizing a figure to being able to write a formal geometric
proof. According to de Villiers [27] the Van Hieles attributed the dismal performance by
students in geometry to the fact that the curriculum was presented at a higher level than
those of the students. As a result the students could not understand the teacher and the
teacher could not understand why the students could not understand.

Young [93] argues that according to Piaget in order for students to gain the ability to
construct a proof, they must progress through certain stages of learning. At stage one a
student is non-reflective, unsystematic and illogical. At this level students will explore
randomly without a plan and will be unable to generalize from one example to the next.
At stage two students will begin to establish relationships, anticipate results and think
logically about premises they believe in. Students at this level are still not ready to
conquer the concept of proof but are beginning, on their own to use some informal
reasoning to justify conclusions. At stage 3 (known as the formal operational stage)
students are capable of formal deductive reasoning and can operate within a mathematical
system. At this stage students begin to see that because a statement is always true implies
that it necessarily must be true. Although the age at which deductive reasoning starts is
disputed educational research has shown that different forms of deductive reasoning can
start as early as the elementary grades as can be seen from the following quote from

8
This research was done by means of a questionnaire that was presented to teachers in the southern EMDC
(Education management District Centre) in Cape Town, South Africa - See appendix J.
 28

 
Stylianides et al [80]: “Even though the findings of existing psychological research do
not specify exact ages at which students master different forms of deductive reasoning,
all forms of deductive reasoning we reviewed  begin to emerge in the early elementary
grades.” However since Piaget’s theory deal with understanding in general while the Van
  we are going to take a closer look at the
Hiele theory deals specifically with geometry
Van Hiele theory.
 
Four important characteristics of the Van Hiele theory as summarized by Usiskin [87] are
as follows: Learners progress through the thought levels in a fixed order. In other words
the learner has to pass through the levels sequentially i.e. the student must first pass
through level one before he/ she can get to level two, etc. At each level of thought that
which was intrinsic in the preceding level becomes extrinsic in the current level. Each
level has its own distinct linguistic symbols and own network of relationships connecting
those symbols. Two persons who reason at different levels cannot understand each other.
The Van Hiele theory distinguishes between five different levels of thought. These are
given by Mason [62] as follows:
level 1 (visualization): student recognizes figures by appearance alone, often by
comparing them to a known prototype. At this level students make decisions based on
perceptions not reasoning.
Level 2 (analysis): students see figures as collections of properties. They can recognize
and name properties of geometric figures, but they do not see relationships between these
properties.
Level 3 (abstraction): students perceive relationships between properties and between
figures. At this level, students can create meaningful definitions and give informal
arguments to justify their reasoning. The role and significance of formal deduction,
however, is not understood and hence proof is not understood.
Level 4 (deduction): students can construct proofs, understand the role of axioms and
definitions and know the meaning of necessary and sufficient conditions. At this level
students should be able to construct proofs such as those typically found in a high school
geometry class.
Level 5 (rigor): students at this level understand the formal aspects of deduction, such as
establishing and comparing mathematical systems. Students at this level can understand
the use of indirect proof and proof by contrapositive and can understand non- Euclidean
systems.

Since the majority of students in our study were Black and Coloured from previously
disadvantaged groups in South Africa it is imperative that we determine what research
says about their proving abilities. It is important to note that the majority of these students
study mathematics in their second language. Research done by de Villiers and
Njisane [26] in 1987 has shown that about 45% of black learners in grade 12 in Kwa-
Zulu Natal had only mastered Van Hiele level 2 or lower, whereas the examination
required mastery at level 3 and beyond. De Villiers quotes research done by Malan
(1986), Smith and De Villiers (1990) and Govender (1995) that shows similar results. He
argues that the transition from level 1 to level 2 is particularly problematic for learners
that have English as a second language since it involves the acquisition of the technical
terminology by which the properties of figures need to be described and explored.
 29

 
De Villiers [29] is of the opinion that improvements in school geometry results are
dependent on major revisions in primary school geometry. He uses the example of a
  Van Hiele theory to prove his point.
Russian experimental curriculum based on the
 
In the late sixties the Russians wanted to determine why students that showed progress in
other subjects performed worse in geometry. They came to the conclusion that the main
 
reason for this state of affairs was insufficient attention to geometry in the primary
school. They discovered that in the first five grades learners dealt mainly with Van Hiele
level 1 activities, while from grade 6 learners suddenly had to deal with activities at a
level 3 understanding. The Russians subsequently developed a very successful
experimental geometry curriculum based on the Van Hiele theory. In this experimental
curriculum they developed the learners’ understanding sequentially and hierarchically
(based on the Van Hiele levels) from grade 1. This had as a result that the average grade
8 of the experimental curriculum showed the same or better geometric understanding than
their grade 11 and 12 counterparts in the old curriculum.

In South Africa until very recently the version of Euclidean geometry that was followed
in high school was one in which congruency and similarity is a central theme. With this
congruency geometry an axiomatic-deductive system of proving was utilized.
De Villiers [29] argues that an axiomatic-deductive approach to teaching is used when an
unfamiliar topic is presented to students by means of the initial introduction of the axioms
and definitions of that topic and logically deriving the other statements (theorems) and
properties from them. This approach however has its limitation and a number of authors
have critiqued the deductive presentation of mathematics. Some like Kline believe that
students may develop feelings of inferiority since the deductive presentation of
mathematics might lead the students to believe that mathematics is created by geniuses
who start with axioms and reason directly and flawlessly to theorems. De Villiers lists the
following as added criticisms of the axiomatic deductive approach to teaching. Learners
cannot interpret work presented in this manner as axiomatic structures and deductive
proof belong to Van Hiele level 4 and most learners enter high school only at the first or
second Van Hiele level. Also no provision is made in such an approach to help learners
progress towards the necessary or required levels. The teaching of axiomatic structures as
finished products lead to the rote learning and memorization of axioms, theorems and
proofs with little or no understanding of their meanings. The students do not become
skilful in the application of creative mathematical processes like abstraction,
generalization, defining, drawing of analogies, systematization, construction of proofs,
etc. since only the end products of these processes are given directly to them.

Since teachers at school play a crucial role in learners’ understanding of proof we have to
determine what their perceptions of proof are. During a 1984 country-wide survey at 11
South African universities De Villiers [26] found that nearly half of the students in
mathematics education had a traditional view of proof i.e. proving based on deduction.
These students therefore believed in an axiomatic deductive approach to teaching proof.

It would seem then based on the above arguments that problems that learners have with
geometry in the lower grades is carried right through the system and contributes
 30

enormously to the problems learners has with  proving. Thus to change this state of affairs
one has to start with the lower grades, making certain that they go through the Van Hiele
 
levels hierarchically and sequentially. This should not only be the case for geometry, but
should include proof in general, where learners should be given experience in proof not
only where proof is done to verify, but also  to explain, to systematize, to discover, to
communicate, etc. This should be done right through all the grades so that learners will
 
eventually become adept at proving. Several researchers agree with this as can be seen
from the following quote from Stylianides et al [80]: “…several researchers have
identified students’ abrupt introduction to proof in high school as a possible explanation
for the many difficulties that secondary school and university students face with proof,
thereby proposing that students engage with proof in a coherent and systematic way
throughout their schooling.”

3.6 The distinction between a priori and a posteriori knowledge

Barker [3] a mathematical philosopher asserts that a distinction that have long been a
topic of discussion in the philosophy of mathematics has been the distinction between a
priori and a posteriori (or empirical) knowledge. He argues that this distinction is
regarded as fundamental in philosophy of mathematics and I would argue that this is also
very important in the teaching and learning of proof in mathematics as this will determine
how one views the attainment of knowledge in proving.

In order to explain empirical knowledge he uses the example of someone that knows
ravens are black. Now in order for the person to know this the person must understand
what is meant by black and what is meant by raven. So the person must have seen ravens
or have heard reports of ravens. In other words the person must have experienced this by
means of their senses which include seeing, hearing, feeling, smelling or tasting. Hence
the point is that only sensory observations can provide the kind of justification needed to
entitle a person to say that he/ she knows facts like these. Therefore a posteriori
knowledge can be defined as that knowledge that requires justification from experience.

To explain a priori let us examine an example of number theory. If a number is an even

number then it is divisible by two; no prime number is divisible by two; therefore no
prime number is even. To come to the conclusion that prime numbers is not even one
does not need experience, but instead used deduction based on the first two premises. So
this is purely a cognitive exercise and knowledge gained in this way is a priori. A priori
knowledge therefore does not depend on experience and hence can be defined as
knowledge that does not need to be justified by experience.

The knowledge of this distinction between a priori and a posteriori is important since it
allows us to see that subjects like physics and biology are concerned with empirical
knowledge since they rely on observations to reach conclusions. In contrast mathematical
logic is concerned only with a priori knowledge since it only seeks knowledge of the
rules governing the validity of an argument and therefore need not rely on observations to
reach conclusions in its arguments. I would therefore argue that since logic and the
majority of proofs are only a priori this is why students find it so difficult to apply and
 31

teachers find it difficult to teach since all   the reasoning takes place cognitively and
therefore one does not know where the reasoning problems are since the reasoning
process is not readily observable.  

3.7  
Cognitive processes involved in learning proof
 
Argument form plays a crucial role in deductive logic. Epp [34] has the following to say
in this regard: “The central concept of deductive logic is the concept of argument form.”
But what constitutes an argument? An argument can be seen as a sequence of statements
aimed at demonstrating the truth of an assertion. Stated differently it means that one can
only be confident in the conclusion that you draw from an argument if the statements
composing it either are acceptable on their own merits or follow from preceding
statements. The following statement from Epp stresses the importance of logic in
mathematical reasoning: “Logical analysis won’t help you determine the intrinsic merit
of an argument‘s content, but will help you analyze an argument’s form to determine
whether the truth of the conclusion follows necessarily from the truth of the preceding
statements. For this reason logic is sometimes defined as the science of necessary
inference or the science of reasoning.” To get a better understanding of argument forms
and deductive reasoning I think it is imperative that we examine the cognitive processes
and mental structures involved in this kind of reasoning.

McNally [66] in his study on Piaget and education use the example of a small child that
sees sheep for the first time. The child exclaims: “Look at the puppy dogs.” The reason
McNally advanced for this erroneous classification is that the child saw the sheep in
terms of that part of his cognitive structure (schema)9 which seemed to apply. What this
means is that the child at this point in his life having encountered four-legged woolly
animals called dogs applied this schema to the sheep, since that was the only schema
available to the child. McNally therefore argues that environmental events such as the
above are assimilated into the cognitive structures not as a mechanistic transaction, but
that the cognitive structures imposes its own organization, meaning or interpretation on
an external stimulus. This means that the child compared the sheep to what was available
in his cognitive structures for small woolly four-legged animals and since only dog was
available interpreted it as dog. So assimilation is the intellectual process whereby the
individual deals with the environment in terms of his present cognitive structures. This
state of affairs however only exists until such time that accommodation in the cognitive
structures of the child occurs. Accommodation occurs when the cognitive structures are
forced to modify by the demands of the environmental event. In other words once the
child have a more complete schema of a dog the child is forced to modify his cognitive
structure in relation to the sheep. McNally states this as follows: “One very important
feature of structures (schemas) is that they change as the result of the interaction of
maturation and experience…”

This example makes it clear then that accommodation cannot proceed without
assimilation and that assimilation and accommodation occur simultaneously. Anderson,

9
Cognitive structures is intellectual structures that the individual has available for the interpretation and
solution of problems posed by the environment.
 32

Reder and Simon [1] are of similar conviction  as can be seen from the following quote:
“A more careful understanding of Piaget would have shown that assimilation of
 
knowledge also plays a critical role in setting the stage for accommodation…that the
accommodation cannot proceed without assimilation.” To emphasize this point let us take
 
an extreme example. One cannot teach a six-year old child, who has just started school,
the concept of a limit in mathematics since the child does not have a cognitive
 
representation available to start an accommodation process from. In other words the child
does not have the requisite mathematical knowledge to make a connection with the
concept of limit. An important question is: Are learners of Euclidean geometry and proof
in general in South Africa calling a sheep a dog because their cognitive structures have
not been developed to maturation because of lack of experience and practice?

Possible answers to this question are given in the previous section by the research done
by the Russians and by the mental models theory which will be discussed later. The
Russians discovered that in the first five grades learners mainly deal with Van Hiele
level 1 activities, while from grade 6 learners suddenly had to deal with activities at a
level 3 understanding. This implies that the learners did not go sequentially and
hierarchically through the reasoning levels as required by Van Hiele. De Villiers [29] has
pointed out that a similar situation exists in South Africa namely that grade 12 learners
have only mastered Van Hiele level 2 or lower, whereas the examination requires mastery
at level 3 and beyond. De Villiers indicated that most learners enter high school at Van
Hiele level 2, but the axiomatic-deductive proofs belong to level 4. He also indicated that
no provision is made to help learners to progress towards the required levels. As a
consequence of the fact that learners did not go through the Van Hiele reasoning levels
sequentially and hierarchically they therefore have not assimilated all the required
knowledge necessary to build a complete cognitive structure required by the
accommodation process. Stylianides and Stylianides is of similar conviction, this can be
seen in the quote that we used earlier namely: “…several researchers have identified
students’ abrupt introduction to proof in high school as a possible explanation for the
many difficulties that secondary school and university students face with proof, thereby
proposing that students engage with proof in a coherent and systematic way throughout
their schooling.” As a result of the above arguments students find proof and proving at
school level a very frustrating and difficult exercise.

The other part of the answer to the above question can be taken from theories concerning
deductive reasoning. A relatively recent psychological theory of deductive reasoning is
the mental models theory. Mental models theory refers to a representation in the mind
that has a structure analogous to the structure of the situation it represents. This theory
was developed by Johnson-Laird and hypothesizes about the process of deductive
reasoning as it is applied to syllogisms10. This theory is described by Stylianides and
Stylianides [80] as follows. The theory consists of three main stages. In the first stage,
known as the comprehension stage, the reasoner constructs a mental model of the
information presented in the premises of a syllogism. In the second stage known as the
description stage the reasoner tries to devise a concise description of the model
constructed in the first stage that concludes something not explicitly stated in the
10
Syllogisms occur when one argues from premises to an inference or a conclusion.
 33

 
premises. The third stage known as the validation stage is where any essential deductive
work is carried out. In this stage the reasoner also searches for counterexamples to the
conclusion drawn in the second stage.  

 
Stylianides et al [80] argue that the effectiveness of this model depends on a person’s
11
working memory capacity. They argue further that limitation in working memory
 
capacity results in errors in reasoning since people fail to consider all possible models of
the premises that would provide them with counterexamples to the conclusions they have
derived from their initial models. They use the example of Euclidean geometry to
illustrate how working memory can be limited by too many separate sources of
information. Sometimes when students are required to do proving in geometry they are
provided with a diagram and a set of given information. To make sense of the two
sources of information (diagrams and the givens) the students must mentally integrate the
information. In other words read the givens, hold them in their working memory and then
search the diagram for the appropriate places to apply it. Stylianides et al argue that this
integration process is cognitively demanding and occupies a significant part of students’
working memory capacity. To overcome errors in reasoning that occur as a result of too
many pieces of separate information having to be processed at the same time, practice is
suggested as a possible solution. Conversely Stylianides et al argue that practice can
however also be associated with secondary aspects of student engagement with proof
such as that of writing proofs in the two-column format. In such proofs the emphasis is
on form and the result is that the proof becomes a ritual procedure and does not enhance
the working memory capacity. The effective use of working memory and consequently
deductive reasoning is therefore dependant on the reasoner freeing up the working
memory by keeping essential information in the long term memory.

3.8 Physiology of learning

Recent discoveries of the workings of the human brain have opened up a host of new
possibilities for teaching and learning. McGeehan [65], for example, has the following
opinion in this regard: “While scientists caution that they are only beginning to unravel
the secrets of how humans learn, what they have already uncovered provides
groundbreaking insights for educational practice. For the first time in the history of
formal schooling, we have the opportunity and challenge to understand and act on the
biology of learning rather than simply following traditional practices.” What McGeehan
means by biology of learning is that we should strive to understand what happens in the
brain when learning takes place. She states this as follows: “Since students’ mastery of
the school curriculum happens primarily in the brain, it stands to reason that educators
should be experts on the workings of that amazing organ.” The question is what happens
in the brain when learning takes place? The learning process in terms of what happens in
the brain is explained by Sylwester [82] as follows: “From a cognitive perspective
learning is explained as the building of neural connections12.” The implications of this for
pedagogy he explains as follows: “A familiarity of basic neural connectivity and brain

11
Working memory is a system for temporarily storing and managing the information required to carry out
complex cognitive tasks such as learning, reasoning, and comprehension.
12
Neural connections are connections between neurons.
 34

structure leads to greater understanding of  how the brain thinks, comprehends and
ultimately learns.” McGeehan [65] further elucidates this, as follows: “New experiences
  13, the brain cells principally involved in
physically change the brain by causing neurons
cognition, to sprout new branches, or dendrites14, and thus increase communication
among neurons across microscopic gaps called  synapses15. The synaptic leap of an
16
electrical impulse between the axon of one neuron and the dendrite of another is the
 
physical basis of learning and memory. When a pathway of communication within a
network of neurons is used repeatedly, it becomes increasingly efficient and we say that
we have learned something.”

13
Neurons are a type of brain cell that receives stimulation from its branches, or dendrites, and
communicates to other neurons by firing a nerve impulse along an axon.
14
Dendrites are branch-like structures that extend from the neuron cell body and receive messages from
other neurons.
15
Synapses are the microscopic gap between the axon of one neuron and the dendrite of another.
16
An axon is the part of a neuron that transfers a nerve impulse from the neuron cell body to a synapse with
another cell.
 35

 
CHAPTER 4

THE INFLUENCE OF EMOTION,   CONFIDENCE, EXPERIENCE AND

PRACTICE ON THE LEARNING PROCESS
 
4.1 Introduction
 
A few years ago while I was teaching a grade 12 mathematics class I discovered that one
of the girl learners, Shahieda17 had major difficulties in learning mathematics. While
teaching the class I would do some examples of a topic on the board and then request that
the learners attempt one or more similar problems on their own. While the learners were
attempting the problems I would walk around in the class and invariably I found that
Shahieda was struggling. I would then sit next to her and point out the mistakes in her
solution and then proceed to correct her argument. Subsequently I would request the class
to attempt another problem of the same kind. Shahieda would then provide the correct
solution to this follow-up problem. Satisfied that she was now cognizant of the
requirements of such problems I would proceed to the next topic.

To my amazement when a test was written that covered the same topic Shahieda would
make exactly the same mistakes that I had pointed out to her. Her results in the two
exams of that year were even worse. She scored less than 10% for both the first and
second paper of both exams. This was despite my numerous attempts to get her to reason
correctly. What puzzled me was that in the class she would seem to understand and even
provide correct arguments, but as soon as she wrote tests or exams she would revert back
to her erroneous arguments. I therefore came to the conclusion that factors other than the
normal reasoning problems were responsible for this behaviour.

4.2 The influence of emotion

An intriguing question is how does the human brain deal with new incoming
information? How the human brain deals with incoming information is explained by
McGeehan [65] as follows: “First the sensory stimuli hit the neurons in the appropriate
sensory cortex18. These crude sensations are then relayed through the thalamus19 and sent
to the sensory association area of the neocortex where they are put together into objects
we recognize. Next (and almost simultaneously) the information is sent to the amygdala20
for emotional evaluation and to the frontal cortex for content evaluation. On the basis of
its analysis of physical features of the stimuli, the brain begins to construct meaning.”

Emotions can affect learning, in both a positive and negative way. Kort et al [55] confirm
this in the following quote: “When a learner experiences positive emotions, the learning

17
Not her real name
18
The cortex is a neuron-packed outer layer of the brain in which conscious thought takes place.
19
The thalamus is a sensory relay station located deep within the middle of the brain.
20
The amygdala is an almond –shaped structure in the middle of the brain, connected to the hippocampus,
which detects the emotional content of sensory data and plays a role in the formation of emotion-laden
memories.
 36

 
process can be enhanced. When a learner experiences negative emotions, the learning
process can be disabled.” Goleman [39] is of similar opinion as the following quote
shows: “Students who are anxious, angry,   or depressed don’t learn; people who are
caught in these states do not take in information efficiently or deal with it well.” The
importance of the effect of emotion on the   learning process is shown in the following
quote from Sylwester [82]: “Emotion drives attention, which drives learning, memory
 
and problem solving and almost everything else we do …by not exploring the role that
emotion plays in learning and memory, our profession has fallen decades behind in
devising useful instructional procedures that incorporate and enhance emotion.”
Sylwester further emphasizes this as follows: “Far more neural fibers project from our
brain’s emotional center into the logical/ rational centers than the reverse, so emotion is
often a more powerful determinant of our behaviour than our brain’s logical/ rational
processes.”

Incoming information also needs to have personal meaning and emotional importance for
students as a prerequisite to being stored in the long term memory. In other words
students tend not to remember for long those things that lack personal meaning and
emotional importance for them. This is confirmed in the following quote of
McGeehan [65]: “…when information lacks personal meaning and an emotional hook,
the neural networks needed to create long-term memories are not formed.” It is clear
therefore that emotion plays a crucial role in the learning process.

When I was at primary school some of the mathematics teachers used to call learners to
the board to solve mathematics problems on the board. The teacher would stand behind
the learner with a cane and as soon as the learner made a mistake the teacher would beat
the learner. Another popular method was where the teacher would do mathematics
revision with a cane. Any learner that could not supply an answer to a question would
routinely receive a beating. I think the teacher’s motivation for using these methods was
that the teacher was under the impression that the learners were lazy and that the beatings
and resulting pain would inspire them to become industrious and would cause them to
learn and understand. The strongest emotions that these learners therefore came to
associate with mathematics is one of paralyzing fear, humiliation and also that not
understanding is a bad thing. The sad part is that neuroscience has shown that memories
that are strongly charged with emotion are most likely to go into the long term memory.
So these learners that had a traumatic experience with mathematics will tend to remember
it for a long time. The result of this is that since all the previous experiences of these
learners with mathematics have been bad they expect that their next experience with the
learning of mathematics would also be bad. This then becomes a self-fulfilling prophecy.
In other words the learner expects a negative experience because the emotions in their
long term memory prepare them for a bad experience and hence they get a bad
experience. Such learners will struggle with mathematics not necessarily because they do
not understand it, but because their previous experience dictates that it must be so. Hence
contrary to their aim of teaching the learners to learn and understand, these teachers
instead taught them anxiety, fear, humiliation and confusion.
 37

Even teachers that do not use intimidation as   a method of teaching and that have good
intentions sometimes do not take into consideration the effect of emotion on the learning
 
process. This is starkly illustrated in the following quote from Kort et al [55]: “When
teachers present material to the class, it is usually in a polished form that omits the
natural steps of making mistakes (e.g. feeling   confused) recovering from them (e.g.
overcoming frustration), deconstructing what went wrong (e.g. not becoming dispirited),
 
and starting over again (with hope and enthusiasm). Those who work in science, math,
engineering, and technology (SMET) as professions know that learning naturally
involves failure and a host of associated affective responses.” Kort et al therefore
conclude that what we fail to teach students is that these feelings associated with various
levels of failure are normal parts of learning, and that they can actually be helpful signals
for how to learn better. It is evident therefore that instructors in mathematics should be
cognizant of the emotional requirements of their subject and should include examples that
allow students to experience a variety of emotions in the learning process so as to prepare
the student for the emotional rigours involved in doing mathematics.

4.3 The influence of Confidence

It is my contention that in order to become proficient in theorem proving (or any

mathematical process) students need to be confident in what they are doing. The literature
is commensurate with my view of the connection between confidence and success in
mathematics. Burton [9] for example states: “This teacher is reflecting a widely held view
that performance in mathematics and confidence go hand-in-hand. Success in
mathematics breeds confidence. Confidence in mathematics breeds success.” Research
done in Canada by PISA21 [10] found that a student’s self-confidence and level of anxiety
about mathematics were strongly associated with their performance. They found that
those with high levels of confidence in their ability to learn mathematics performed much
higher than those with low levels. Conversely students with a high level of anxiety about
mathematics, such as feelings of helplessness or stress when dealing with mathematics,
performed much lower than students with less anxiety. Clute [15] is of the same opinion
as can be seen from the following quote: “Hence, people high in mathematics confidence
perform better on mathematical tasks. Mathematics anxiety is strongly but negatively
related to mathematics confidence.”

In order to inform our discussion on the relationship between confidence and success in
mathematics, I think it is imperative that we define what we mean when we are referring
to confidence. Although there are different views as to the meaning and definition of
confidence in mathematics, our understanding of mathematical confidence will be
informed by that of Burton [9] which is as follows: “…I saw confidence as a label for a
confluence of feelings relating to beliefs about the self and about one’s efficacy to act
within a social setting, in this case the mathematics classroom.” Since the two main
protagonists in the mathematics classroom are the teacher and the student we therefore

21
PISA (Programme for International Student Assessment) is a collaborative effort among member
countries of the Organisation for Economic Co-operation and Development (OECD) – this program is
designed to regularly assess the achievement of 15-year-olds in reading, mathematical and scientific
literacy using a common international test.
 38

have to ascertain their views on confidence.  The study done by Burton found that
teachers regarded confidence as individual and behavioural. That is the teachers did not
 
think that confidence in mathematics involved a social act and that confidence in the
individual is exhibited by behaviour such as willingness to answer or to attempt.
  with feelings and how the classroom could
Conversely the students associated confidence
function to make those feelings better or worse. The students were in favour of a
 
collaborative working style and were of the opinion that getting answers correct fueled
the confidence level. Furthermore they were convinced that both knowledge and
understanding contributed to confidence.

If confidence is as crucial as the literature suggests an important question then would be

what do teachers have to do in order to inculcate their students with confidence in
mathematics? I am of the opinion that some of the student responses of the Burton [9]
study provide an answer to this question. These include the following: “They wanted
teachers to facilitate discussion, teamwork, a light-hearted approach, a relaxed classroom
where you are not afraid of making errors…they did not want to be put down, persistently
asked the same question, made to look a fool or feel patronized, put into a position where
others laugh at you, …the students felt that teachers should explain well, should not rush
the work, should know what they are talking about and should be sensitive to students
who are struggling to understand.” Clute [15] on the other hand was of the opinion that
instructional strategy plays a crucial role in mathematics achievement. The study done by
Clute indicates that students with low levels of anxiety and hence high confidence were
better served by the discovery method of teaching22. Conversely students with a high
level of mathematical anxiety and low confidence relied heavily on a well-structured,
controlled plan for learning.

4.4 The influence of Experience

Another question that deserves our attention is: Is there a relationship between
confidence and experience? Anderson, Reder and Simon [1] provide the following
answer to this question: “Cognitive competence (in this case mathematical competence)
depends on the availability of symbolic structures (e.g. mental patterns or mental images)
that are created in response to experience.” The answer provided by neuroscience is much
more compelling. Findings from brain research indicate that intelligence is a function of
experience. McGeehan [65] explains it as follows: “New experiences physically change
the brain by causing neurons, the brain cells principally involved in cognition, to sprout
new branches, or dendrites, and thus increase communication among neurons across
microscopic gaps called synapses…The findings of neuroscientists affirm the importance
of experience in the development of dendrites and, by extension, in the results of this
development which we call learning and observe as intelligence.” Clute [15] has the
following opinion about the relationship between confidence and the learning of
mathematics: “If one lacks confidence in one’s ability to perform mathematical tasks, it
seems reasonable to conclude that there is a lack of respect for or trust in one’s own

22
The discovery method of teaching was devised by Clute for her study. In this method the focus is on the
teacher interacting with students to develop subject matter concepts from which the students then
“discover” the answer.
 39

instincts or judgments when it comes to  learning mathematics.” Therefore since

confidence influences the learning of mathematics and in turn learning is influenced by
 
experience it is clear that experience does influence confidence. I would therefore suggest
that the more experienced one becomes in mathematical procedures and techniques the
more your confidence will be enhanced. An  important next question then, is how does
one gain experience in mathematics?
 
4.5 The influence of Practice

Bransford et al [7] uses chess to provide an answer to the above question. They argue that
it is estimated that world-class chess masters require from 50 000 to 100 000 hours of
practice to reach that level of expertise. Much of the practice time involves the
development of pattern recognition skills that support the fluent identification of
meaningful patterns of information plus knowledge of their implications for future
outcomes. Bransford therefore quote Singley and Anderson to make the point that in all
domains of learning, the development of expertise occurs only with major investments of
time, and the amount of time it takes to learn material is roughly proportional to the
amount of material being learned. There is a belief among many people that talent
determines who becomes an expert in a particular area of learning. Ericsson et al [35]
however have found that even seemingly talented individuals require a great deal of
practice in order to develop their expertise. It is clear therefore that in order to gain
experience one has to do extensive practice. Anderson, Reder and Simon [1] states for
example that the last 20 years of research on cognitive psychology showed that real
competence only comes with extensive practice. They state: “The instructional task is not
to “kill” motivation by demanding drill, but to find tasks that provide practice while at the
same time sustaining interest.” Vygotsky [89] was of similar opinion and writes in this
regard: “Wundt long ago established that the latent period of a complex reaction
decreases with practice.”

4.6 Negotiating the learning process

Based on the above arguments the following is our view as to some of the things that
contribute towards making a student successful in the study of mathematics. First the
teacher has to endeavor to create an atmosphere in the classroom that students find non-
threatening. In order to become confident, the student needs to practice what he /she has
been taught. This is the way it works: the student in the class has his/her zone of proximal
development (zpd)23 extended by his/her teacher or more capable peer. In other words the
student is helped by the teacher or peer to do problems that the student could not master
on his/her own. This is because according to Vygotsky [89] what is in the zone of
proximal development today will be the actual developmental level tomorrow. In order to
assimilate this (i.e. to make it part of his/her knowledge structures) the student now needs
to practice. This has to be done by the student on his/her own. This also requires the
student to invest a lot of time into practice. MGeehan [65] confirms this as follows:

23
The zone of proximal development is the distance between the actual development level as determined
by independent problem solving and the level of potential development as determined through problem
solving under adult guidance or in collaboration with more capable peers.
 40

“When a pathway of communication within  a network of neurons is used repeatedly, it

becomes increasingly efficient and we say that we have learned something.”
 
What this also requires is a commitment on the side of the student which requires the
 
student to have a thirst for this kind of knowledge i.e. there must be another reason
besides studying for tests or exams to master the required concepts in mathematics. That
 
is the student must have an emotional connection with the knowledge i.e. a hunger for
knowledge for knowledge’s sake. This then becomes a cycle i.e. a student is extended
beyond his/her zone of proximal development then he/she consolidates this by practicing
and hence become more confident as he/she masters the subject matter. This confidence
therefore is fuelled by being successful in attempting the subject matter and hence
building more experience.

Assessment will help the teacher identify which reasoning problems students still have.
So after assessing students the teacher should try and identify, based on their mistakes,
where the students still have gaps or faults in their reasoning. The teacher should then
devise strategies to help the student in overcoming these problems with their reasoning.
Once the teacher is satisfied that the student is now reasoning correctly, the student
should be reassessed to confirm that there are no more gaps or faults in their reasoning.
The fact that the student will now be able to successfully solve these kinds of problems
will further boost his/her confidence and at the same time provide the student with
valuable experience.
 41

 
CHAPTER 5

 
RESEARCH METHODOLOGY

5.1 Introduction  

 
The students of the MAM 112 course (a first year elective mathematics course) at UWC
were utilized by us as the experimental group. At the start of the semester 39 students
were enrolled for this course. For various reasons seven of these students dropped out.
Six of the remaining 32 were students that repeated the course and therefore were not
taken into consideration for the statistical analysis of the study. The students of the MAM
111 course (a first year core module at UWC) formed the control group of the study.
Only 21 of these students completed both the pre and post-test of the logic unit and
therefore only the results of these students were taken into consideration for statistical
analysis.

Our intention was to complete the study during the first semester of 2007. However we
could not finish all the content in the first semester and therefore had to use 3 to 4 weeks
of the second semester as well. The total amount of learning hours available was 150
hours, 96 hours of which was utilized as contact time with students. The contact sessions
consisted of four one hour lectures and a two hour tutorial session per week. The
experimental group received teaching on logic and thereafter teaching on the different
types of proof whereas the control group received no teaching on logic but were taught
on topics of Differential Calculus. The textbook that was used by the experimental group
was “Discrete Mathematics with Applications” by Susanna S. Epp [33].

The study used a pre-test post-test design. The measuring instruments for this study were
mainly these pre- and post-tests. Only for the first logic puzzle an experimental-control
group design was used. For all the logic puzzles the pre-test was given without the
students having received any teaching on the given topic. The students received no
feedback on the pre-test. After the pre-test was administered, the experimental group
was taught the logic skills involved in the puzzles. A post-test was then administered to
determine if students had acquired the requisite skill. The experimental group of students
also wrote class tests that counted towards their course mark but these tests were not
utilized as a measuring instrument in the study. Conversely the pre- and post-tests did not
count towards the course mark, but formed part of the measuring instruments of the
study. The tests were administered during the regular class meetings. Although both
quantitative and a qualitative designs were utilized in the study, the majority of the
findings were based on a quantitative design.

During their very first lecture both the control and experimental groups were asked to
solve a logic puzzle of the Lewis-Carol type. At this stage none of the students had
received any formal instruction in mathematical logic. The reason why we gave the
students pre-tests on the logic puzzles without first teaching them is based on an
experiment of Piaget. In this experiment Piaget asks a five-year –old child why the sun
does not fall. What he is assuming, according to Vygotsky [89], is that the child does not
 42

have a ready answer for such a question nor  the capabilities to generate one. Furthermore
the point of asking a question that is so far beyond the reach of the child’s intellectual
skills is to eliminate the influence of previous  experience and knowledge. So the aim is to
obtain the tendencies of children’s thinking in “pure form” i.e. thinking that is entirely
 
independent of learning. This pure form of thinking is what we wanted to determine of
our students i.e. what their current cognitive abilities were in terms of deductive
 
reasoning.

5.2 Learning theory

We support the information-processing learning theory of constructivism. This

information –processing approach is based on an approach espoused by Anderson, Reder
and Simon [1]. According to this approach learning is viewed as an active, constructive
process in which students attempt to resolve problems that arise as they participate in the
mathematical practices of the classroom. Our teaching therefore was designed to embrace
this kind approach.

5.3 Teaching Methodology

The teaching methodology was discussion-based with problems discussed and solved
during lecture periods. Typically the lecturer would do some examples on the board,
which would then be discussed. The discussion normally was preceded by the lecturer
asking pertinent questions to determine if the students understood the examples. If it
became clear that students did not understand the examples the lecturer would then
initiate a discussion based on student questions. Once we were satisfied that the majority
of students were aware of the cognitive requirements of the examples we posed problems
similar to the examples to students. These problems would then be attempted by students
and if necessary with help from more capable peers, the teaching assistant or the lecturer.
Sometimes the lecturer would ask one of the students to do his/her solution on the board.
The student was expected to not only write down the solution on the board, but also to
explain his/her solution. These explanations gave us insight into the students’
understanding and reasoning. These solutions were also discussed and gaps or errors in
reasoning were pointed out and corrected immediately. This was done in order to
consolidate what the students have learned. This consolidation exercises were given
immediately after the lesson and was mostly done in class. The importance of this
approach is emphasized by Kutzler [58] in the following quote: “In the psychology of
learning, scientists discovered the concept of reinforcement and showed that
reinforcement works best if it follows the action immediately”. At the end of the lecture
students would be given exercises to complete as homework. The solutions to the
homework were provided in the subsequent lecture.

5.4 The logic component of the course

As already indicated the first logic puzzle was given to the students in the very first
lecture, before they had any exposure to teaching on the subject of logic and therefore
this puzzle was used to test the initial deductive abilities of students. This puzzle was
 43

named puzzle I and was given as the first  pre-test. Puzzle I contained compound and
conditional statements. The puzzle is presented in Appendix A1.
 
The following day the students were given another puzzle, i.e. puzzle 2. This puzzle
 
known as the knights and knaves puzzle contained only two statements and a question.
The solution to the puzzle was based on contradiction. The students were given this
 
puzzle without prior teaching on contradiction and therefore this puzzle formed
pre-test 2. This pre-test was used to ascertain the students’ level of comprehension of the
use of contradiction as a method of proof. No immediate feedback was provided to
students as to the solution of the puzzle. The puzzle is presented in Appendix B1.
Students were required to write their names on the puzzles so that we could track each
student individually.

Both the experimental and control group students received no feedback on puzzle I. In
the subsequent lectures the experimental group received teaching on logic. We started
with the logic of compound statements. In propositional logic the words sentence, true
and false are the initial undefined terms. We then proceeded to definitions of statements
and compound statements and the symbolism involved in these and thereafter to truth
tables. This was followed by an investigation of the logical equivalence of statements
which in turn was followed by conditional statements. Conditional statements were
followed by discussions on valid and invalid arguments. The logic component of the
course was ended with teaching on the logic of quantified statements.

5.4.1 Compound Statements

Statements are defined as a sentence that is either true or false but not both. A compound
statement is formed when statements are joined by the logical connectives: and ( ∧ ) ,
or ( ∨ ) and not ( ~ ).

5.4.1.1 Truth tables

The truth table for a given statement form displays the truth values that correspond to the
different combinations of truth values for the statement variables. For a statement to have
well-defined truth-values means that it must either be true or false, but not both. So if p is
a statement then the negation of p is “not p” and in symbol form is given as “~ p”. This
means that if p is true then ~ p is false and conversely if p is false then ~ p is true. The
truth table is shown in table 1 below (T = true and F = false).

Truth table for ~ p

p ~p
T F
F T
Table 1
 44

The conjunction of statements p and q is “p and  q” denoted p ∧ q . It is true when and only
when both p and q are true. If either p or q is false, or if both are false, p ∧ q is false. The
 
truth table is shown in table 2 below.
 
Truth table for p ∧ q .
p q   p∧q
T T T
T F F
F T F
F F F
Table 2

The disjunction of statement variables p and q is “p or q” denoted p ∨ q . It is true when at

least one of p or q is true and is false only when both p and q are false. The truth table is
shown below in table 3.

Truth table for p ∨ q .

p q p∨q
T T T
T F T
F T T
F F F
Table 3

To evaluate the truth values of more general statements the following steps must be
adhered to. First evaluate the expressions within the innermost parentheses, then evaluate
the expressions within the next innermost set of parentheses, and so forth until you have
the truth values for the complete expression. An example of such a statement is:
( p ∨ q ) ∧ ~ ( p ∧ q ) . The truth table for this statement is given in table 4 below.

Truth table for ( p ∨ q ) ∧ ~ ( p ∧ q )

p q p∨q p∧q ~(p∧q) (p∨q) ∧ ~( p∧q)
T T T T F F
T F T F T T
F T T F T T
F F F F T F
Table 4

The logical equivalence of statements was investigated next.

 45

5.4.1.2 Logical equivalence of statements  

 
Two statement forms are called logically equivalent if, and only if, they have identical
truth values for each possible substitution of statements for their statement variables. The
logical equivalence of statement forms A and  B is denoted by A ≡ B .
A tautology is a statement form that is always   true regardless of the truth values of the
individual statements substituted for its statement variables.
A contradiction is a statement form that is always false regardless of the truth values of
the individual statements substituted for its statement variables. The statement form
p ∨ ~ p is a tautology and p ∧ ~ p is a contradiction.

The following is a summary of the logical equivalences that can be used to simplify
statement forms: Given any statement variables p, q and r a contradiction c and a
tautology t, the following logical equivalences hold:
1. Commutative laws: p ∧ q ≡ q ∧ p and p ∨ q ≡ q ∨ p
2. Associative laws: ( p ∧ q ) ∧ r ≡ p ∧ (q ∧ r ) and
( p ∨ q) ∨ r ≡ p ∨ (q ∨ r )
3. Distributive laws: p ∧ (q ∨ r ) ≡ ( p ∧ q) ∨ ( p ∧ r ) and
p ∨ (q ∧ r ) ≡ ( p ∨ q) ∧ ( p ∨ r )
4. Identity laws: p ∧ t ≡ p and p ∨ c ≡ p
5. Negation laws: p ∨ ~ p ≡ t and p ∧ ~ p ≡ c
6. Double negative laws: ~ (~ p) ≡ p
7. Idempotent laws: p ∧ p ≡ p and p ∨ p ≡ p
8. De Morgan’s laws: ~ ( p ∧ q) ≡ ~ p ∨ ~ q and ~ ( p ∨ q ) ≡ ~ p ∧ ~ q
9. Universal bound laws: p ∨ t ≡ t and p ∧ c ≡ c
10. Absorption laws: p ∨ ( p ∧ q) ≡ p and p ∧ ( p ∨ q ) ≡ p
11. Negation of t and c: ~ t ≡ c and ~ c ≡ t

The following is an example where the above laws are used to verify a logical
equivalence: ~ ((~ p ∧ q) ∨ (~ p ∧ ~ q)) ∨ ( p ∧ q) ≡ p

Solution
~ ((~ p ∧ q ) ∨ (~ p ∧ ~ q)) ∨ ( p ∧ q)
≡ ~ (~ p ∧ (q ∨ ~ q )) ∨ ( p ∧ q) - by the distributive law
≡ ~ (~ p ∧ t ) ∨ ( p ∧ q) - by the negation law
≡ ~ (~ p) ∨ ( p ∧ q) - by the identity law
≡ p ∨ ( p ∧ q) - by the double negative law
≡ p - by the absorption law

Upon the completion of this section a formal test was given which covered the topics of
compound statements, negation of a statement, conjunction and disjunction, truth tables
and logical equivalences. This test counted towards the course mark, but was not used as
a measuring instrument.
 46

5.4.2 Conditional Statements  

A conditional statement is of the form if p   then q, where p and q are statements (p is

called the hypothesis and q is called the conclusion). This is based on the fact that when
you make a logical inference or deduction you   reason from a hypothesis to a conclusion.
Sentences such as these are called conditional because the truth of statement q is
 
dependent on the truth of statement p. If p then q is denoted by p → q . It is false when p
is true and q is false, otherwise it is true. The truth table for the conditional statement is
shown in table 5.

Truth table for p → q .

p q p→q
T T T
T F F
F T T
F F T
Table 5

In expressions that include → as well as other logical operators such as ∧, ∨ and ~ the
order of operations is that → is performed last.

The negation of “if p then q” is logically equivalent to “p and not q”. Students were
provided with exercises where they were required to symbolically show equivalence as
well as instances where they were required to rewrite the negations of if-then statements
in everyday English. This was done throughout the logic component and was done to
prevent students getting too mechanistic in their reasoning when applying rules of logic.

5.4.2.1 Contra-positive, Converse and Inverse of a conditional statement

A conditional statement is logically equivalent to its contra-positive. The contra-positive

of a conditional statement of the form “if p then q” is “if ~q then ~p”. The fact that a
conditional statement is equivalent to its contra-positive is one of the most fundamental
laws of logic.
The converse of “if p then q” is “if q then p”. The inverse is “if ~p then ~q”. Both the
converse and inverse of a conditional statement is not logically equivalent to the
statement.

5.4.2.2 Bi-conditional

Given statement variables p and q, the bi-conditional of p and q is “p if, and only if, q”
and is denoted p ↔ q . It is true if both p and q have the same truth values and is false if
p and q have opposite truth values. The truth values for the bi-conditional is shown in
table 6 below.
 47

Truth table for p ↔ q .
p q p↔q
 
T T T
T F  F
F T F
F F  T
Table 6

The hierarchy of operations for the five logical connectives are:

1. ~
2. ∧, ∨
3. →, ↔

5.4.2.3 Necessary and sufficient conditions

If r and s are statements, then r is a sufficient condition for s means “if r then s”. On the
other hand r is a necessary condition for s means “if not r then not s”

Upon completion of this unit the students were tested on negations, converse, inverse,
contra-positive, necessary and sufficient conditions and rewriting in if-then form. This
test contributed to their course mark, but was not used as part of the measuring
instrument.

5.4.3 Valid and invalid argument forms

An argument is a sequence of statements. All statements except the final are called
premises or hypotheses. The final statement is known as the conclusion. The
symbol ∴ , read “therefore” is placed just before the conclusion.

What does it mean to say that an argument form is valid? To say that an argument form is
valid means that no matter what particular statements are substituted for the statement
variables in its premises, if the resulting premises are all true, then the conclusion is also
true. When an argument is valid and its premises are true, the truth of the conclusion is
said to be inferred or deduced from the truth of the premises. Truth tables can be utilized
to test the validity of an argument. In order to do this one has to construct a truth table
that shows the truth values of all the premises and the conclusion. In each critical row
(rows in which all the premises are true) determine whether the conclusion of the
argument is also true. If in each critical row the conclusion is also true, then the argument
form is valid. If there is at least one critical row in which the conclusion is false the
argument form is invalid. The following is an example where a truth table is used to
determine if an argument is valid: p ∨ (q ∨ r )
~r
∴ p∨q

The truth table is shown in table 7

 48

Truth  table
p q r q∨r p ∨ (q ∨ r ) ~r p∨q
 
T T T T T F T
T T F T T  T T
T F T T T F T
T F F F T  T T
F T T T T F T
F T F T T T T
F F T T T F F
F F F F F T F
Table 7

The rows that are highlighted are the critical rows. Since in each row where the premises
are both true (critical rows) the conclusion is also true, we conclude that the argument is
valid.

5.4.3.1 Rules of inference

There are rules in logic which state that certain arguments are valid. These rules are
known as the rules of inference. The rules are given below:

1. One of the best known rules of inference is modus ponens. Modus ponens in
Latin means method of affirming. Its argument form is as follows:
If p then q
p
∴q

2. Modus Tollens means method of denying. Its argument form is:

If p then q
~q
∴~ p

3. Disjunctive addition is used for generalization and has the following valid
argument form:
p or q
∴p∨ q ∴p∨ q

4. Conjunctive simplification is used for particularization and has the following

argument form:
p∧q or p∧q
∴p ∴q
 49

5.  
Disjunctive syllogism is the valid argument form that says when you have only
two possibilities and you can rule out one, the other must be the case. Its
argument form is:  
p∨q or p∨q
 
~q ~p
∴p ∴q
 

6. Many arguments in mathematics contain chains of if-then statements. If one

statement implies a second and the second implies the third, then you can
conclude that the first statement implies the third. This valid argument form is
known as hypothetical syllogism. Its symbolic argument form
is:
p→q
q→r
∴p→r

7. It often happens that one knows that one thing or another is true. If you can show
that in either case a certain conclusion follows, then this conclusion must also be
true. This is known as Dilemma: proof by division into cases. Its valid argument
form is:
p∨q
p→r
q→r
∴r
8. Rule of contradiction – suppose p is some statement whose truth you wish to
deduce. If you can show that the supposition that statement p is false leads
logically to a contradiction, then you can conclude that p is true. The
contradiction rule forms the logical base of the method of proof by contradiction.
Its argument form is as follows:
~ p → c [ where c is a contradiction]
∴p
5.4.3.2 Fallacies

A fallacy is an error in reasoning that results in an invalid argument. Two types of

fallacies that were discussed were the converse and inverse error. These are arguments
that resemble modus ponens and modus tollens but are invalid.

The general form of the converse error is as follows:

p→q
q
∴p
This argument form is invalid since a conditional statement is not logically equivalent to
its converse.
 50

The form of the inverse error is as follows:  

p→q
~p  
∴~q  
This argument form is invalid because a conditional statement is not logically equivalent
to its inverse.  

Students were also made aware of valid arguments that have false conclusions and
invalid arguments that have true conclusions.

5.4.3.3 Examples based on argument forms

Upon the completion of the section on valid and invalid arguments students were
provided with a list of the rules of inference. They were then required to deduce
conclusions from given premises providing reasons for each deduction. The following is
an example of such an exercise:

Example 1
Use the rules of inference to deduce the conclusion from the premises, giving a reason for
each step:
1. ~ p∨q → r
2. s∨~q
3. ~t
4. p→t
5. ~ p ∧ r → ~s
∴~q

Solution
6. ~ p from 3 & 4 – modus tollens
7. ~ p∨ q from 6 – disjunctive addition
8. r from 1 & 7 – modus ponens
9. ~ p∧ r from 6 & 8 – conjunctive addition
10. ~s from 5 & 9 – modus ponens
11. ~q from 10 & 2 – disjunctive syllogism

Students were also provided with exercises where they were required to rewrite
statements in symbol form and then use the rules of inference to deduce a conclusion.
The following is an example of such an exercise:

Example 2
You are about to leave for school in the morning and discover you don’t have your
glasses. You know the following statements are true:
(a) If my glasses are on the kitchen table, then I saw them at breakfast.
(b) I was reading the newspaper in the living room or I was reading the
newspaper in the kitchen.
 51

(c) If I was reading the newspaper in  the living room, then my glasses are on the
coffee table.
(d)  
I did not see my glasses at breakfast.
(e) If I was reading my book in bed, then my glasses are on the bed table.
(f) If I was reading the newspaper  in the kitchen, then my glasses are on the
kitchen table.
 
Where are the glasses?

Solution
Let p = my glasses are on the kitchen table
q = I saw my glasses at breakfast
r = I was reading the newspaper in the living room
s = I was reading the newspaper in the kitchen
t = my glasses are on the coffee table
u = I was reading my book in bed
v = my glasses are on the bed table

The statements (a) to (f) then translate as follows:

(a) p→q
(b) r∨s
(c) r→t
(d) ~q
(e) u→v
(f) s→ p

Based on this the following deductions can now be made:

1. ~p from (a) & (d) – modus tollens
2. ~s from (f) & 1 – modus tollens
3. r from (b) & 2 – disjunctive syllogism
4. t from (c) & 3 – modus ponens
Hence t is true and the glasses are on the coffee table.

5.4.3.4 Applying the Contradiction Rule

Students were also provided with exercises where they were required to apply the rule of
contradiction. The students were given such a puzzle as a pre-test i.e. where they were
required to solve a knights and knave puzzle without prior teaching on the rule of
contradiction. This pre-test therefore was used to ascertain the students’ level of
comprehension of the use of contradiction as a method of proof. A tutorial session was
utilized to discuss the solution to this puzzle. After discussing the solution to the puzzle
students were presented with an equivalent puzzle that was used as a discussion exercise.
This was done to illustrate the cognitive processes involved in getting to a solution with
this type of problem. Students were then given a post-test24 to determine if they had
progressed in their deductive abilities where method of contradiction is concerned. The

24
See Appendix B2.
 52

results of the post-test however showed that  approximately 50% of the students still had
problems with the concept of contradiction. Consequently we decided that an intervention
  was utilized for this purpose.
session was necessary. Another tutorial session
  a crucial role in student understanding of
I am of the opinion that teaching strategy plays
the application of certain tools in mathematics. An example of this is the teaching
 
strategy one can employ to connect the rule of contradiction to the method of proof by
contradiction. Our aim was to use the knights and knaves puzzles as a precursor to proof
by contradiction.

Epp [33] states the following about proof: “Probably the most important reason for
requiring proof in mathematics is that writing proof forces us to become aware of
weaknesses in our arguments and in the unconscious assumptions we have made.” The
teaching strategy employed by the lecturer to eliminate reasoning errors of students in
proof by contradiction, is illustrated in the following succinctly described teaching
session. The students were supplied with the following knights and knaves puzzle which
they had to solve:

The logician Raymond Smullyan describes an island containing two types of people:
knights who always tell the truth and knaves who always lie. You visit the island and are
approached by two natives who speak to you as follows:
A says: Both of us are knights
B says: A is a knave.
What are A and B?

The lecturer called Marsha to present her answer on the board.

5.4.3.5 Marsha’s solution:

Suppose A is a knight, then what he says is true, then they both are knights, but B says
that A is a knave ∴ it’s a contradiction because A cannot be both a knight and a knave
∴ A is a knave and B is a knight.

The lecturer then asked the class if they were satisfied with this answer. (Some of the
students’ answers he ignored since these answers were not helping to improve their
understanding of contradiction). Thulani stated that he does not see how Marsha got to
the conclusion that B is a knight. Siphokazi stated that the conclusion and argument do
not connect. Using these two answers the lecturer then proceeded to make certain that the
whole class understood that Marsha did not show how she got from “A and B cannot be
both a knight and a knave” to “∴ A is a knave and B is a knight”. The lecturer also
explained that Marsha’s answer was one long sentence and that one is required to make a
supposition and then make a full stop. One then has to think about what the supposition
implies. The answer to this forms the next statement in the argument. The next statement
is then analyzed for its implications, etc. The aim therefore was to get the students to
reason one step at a time. Using the above argument and input from the students the
 53

lecturer then rectified Marsha’s answer on  the board so that it reflected the correct
answer.
 
5.4.3.6 Marsha’s corrected answer
 
1. Suppose A is a knight.
 
2. Then what A says is true – by definition, of knights
3. Therefore B is a knight
4. Therefore what B says is true – by definition of knights
5. Then A is a knave
6. So A is both a knave and a knight
7. This is a contradiction
8. Hence our supposition is false
9. ∴ A is a knave – negation of supposition
10. ∴ What B says is true
11. ∴ B is a knight – by definition of knight.

The lecturer then asked the class to explain the structure of proof by contradiction. The
following is Earl’s explanation:
“You first make a supposition, which will have consequences. These consequences will
lead to a contradiction which proves the supposition false. Which makes the opposite of
the supposition true.”

After this and other encouraging responses from the students we were satisfied that the
majority of the students now had a reasonable good grasp of the reasoning involved in
proof by contradiction and that we could now proceed to the next topic of discussion.

5.4.4 Quantified statements

Analysis of compound statements can elucidate many aspects of reasoning, but there are
many cases in everyday life and mathematics where it cannot be used to determine
validity. Hence a different methodology is needed to deal with such cases. The logic of
quantified statements provides a means to deal with arguments that cannot be analyzed as
before.

Arguments where sentences need to be separated into parts and where words such as “all”
or “some” play a special role are classified as quantified statements in logic. In normal
English grammar declarative sentences can be separated into subjects and predicates.
Here predicate refers to the part of the sentence that gives information about the subject.
In logic, however a predicate is a sentence that contains a finite number of variables and
becomes a statement when specific values are substituted for the variables. The domain
of a predicate variable is the set of all values that may be substituted in place of the
variable. The set of all elements that make the predicate true is called the truth set of the
predicate. The truth set of a predicate P(x) is denoted by: {x∈ D / P ( x)}
 54

  such as “some” or “all” and tell for how

Quantifiers are words that refer to quantities
many elements a given predicate is true. The symbol ∀ denotes “for all” and is called the
universal quantifier.  

Let Q(x) be a predicate and D the domain of  x. A universal statement is a statement of the
form “ ∀ x ∈ D, Q( x) ” It is defined to be true if, and only if, Q(x) is true for every x in D.
 
It is defined to be false if, and only if, Q(x) is false for at least one x in D. After the above
theory on universal quantifiers was discussed students were given exercises that
contained mathematical examples of applications of the universal quantifier.

The existential quantifier is denoted by the symbol ∃ and means “there exists”. Let Q(x)
be a predicate and D the domain of x. An existential statement is a statement of the form
“ ∃ x ∈ D such that Q( x) ” It is defined to be true if, and only if, Q(x) is true for at least
one x in D. It is false if, and only if, Q(x) is false for all x in D. After the existential
quantifiers were discussed students were presented with exercises that dealt with
mathematical examples of existential statements.

Since it is important to be able to translate from formal into informal language when
trying to make sense of mathematical concepts we provided the students with exercises
that required them to translate universal statements from formal into informal language
and vice versa.

Since a great many of mathematical statements are universal conditional statements of the
form: ∀ x, if P( x) then Q( x) , it is imperative that students are familiar with statements of
this form. We therefore required students to translate universal conditional statements
from formal to informal language and vice versa.

Equivalent forms of universal and existential statements were then tackled. The negation
of a universal statement “all are” is logically equivalent to an existential statement “some
are not”. In formal language it can be given as follows:
~ (∀ x ∈ D, Q( x) ) ≡ ∃ x ∈ D such that ~ Q( x)
Conversely the negation of an existential statement “some are” is logically equivalent to a
universal statement “all are not”. Symbolically it can be given as follows:
~ (∃ x ∈ D, such that Q( x) ) ≡ ∀ x ∈ D, ~ Q( x)
Examples of both forms of negation using both formal and informal cases were discussed
and analyzed in class.

To determine the understanding of the students regarding quantified statements a test was
administered. This formal test counted towards their course mark, but was not used as a
measuring instrument. Since student results were encouraging we continued with other
aspects of quantified statements.

There are many statements in mathematics that contain more than one quantifier. Hence
we examined formal and informal statements that contained multiple quantifiers. We then
proceeded to negations of multiple quantified statements. The following two
 55

generalizations were used to do a number of  formal and informal examples of multiple
quantified statements:
The negation of ∀ x, ∃ y such that  P( x, y ) is logically equivalent to
∃ x such that ∀ y, ~ P( x, y )
 
The negation of ∃ x such that ∀ y P( x, y ) is logically equivalent to
∀ x, ∃ y such that ~ P( x, y )  

5.4.4.1 The Contra-positive, Converse and Inverse of Universal Conditional

statements

Given a statement of the form: ∀ x ∈ D, if P( x) then Q( x) ,

Then its contra-positive is the statement: ∀ x ∈ D, if ~ Q( x) then ~P( x)
Its converse is the statement : ∀ x ∈ D, if Q( x) then P( x) and
Its inverse is the statement: ∀ x ∈ D, if ~ P( x) then ~ Q( x)

The following is an example where the contra-positive, converse and inverse were
applied:

Example 3
Rewrite the following informal sentence in formal language and then give its contra-
positive, converse and inverse:
If a real number is greater than 2, then its square is greater than 4.

Solution
Formal statement: ∀ x ∈ ℜ, if x 〉 2, then x 2 〉 4
Contra-positive: ∀ x ∈ ℜ , if x 2 ≤ 4 then x ≤ 2 [if the square of a real number is less
than or equal to 4, then the number is less than or equal to 2]
Converse: ∀ x ∈ ℜ , if x 2 〉 4, then x 〉 2 [if the square of a real number is greater than 4,
then the number is greater than 2]
Inverse: ∀ x ∈ ℜ , if x ≤ 2 then x 2 ≤ 4 [if a real number is less than or equal to 2, then the
square of the number is less than or equal to 4]

It was shown previously that a conditional statement is logically equivalent to its contra-
positive and that it is not logically equivalent to either its converse or its inverse. The
same is true for universal conditional statements i.e.:
∀ x ∈ D, if P( x) then Q( x) ≡ ∀ x ∈ D, if ~ Q( x) then ~ P( x) and

and

The logical equivalence of a universal conditional statement to its contra-positive is a

very useful tool when dealing with complex statements.
 56

 
It was shown that necessary, sufficient and only if can be extended to apply to universal
conditional statements. Some examples were dealt with where this was illustrated.
Extensive exercises were done where students  were required to write the contra-positive,
converse, inverse and negations of both formal and informal universal statements.
 
5.4.4.2 Universal Instantiation, Universal Modus Ponens and Universal Modus
 
Tollens

The rule of universal instantiation states that if some property is true of everything in a
domain, then it is true of any particular thing in the domain. The rule of universal
instantiation can be combined with modus ponens to obtain the rule called universal
modus ponens. The valid argument form of universal modus ponens can be represented
as follows:
∀ x, if P( x) then Q( x)
P (a ) for a particular a
∴ Q(a)
The argument form of universal modus ponens consists of two premises and a conclusion
and where at least one premise is quantified. The first premise is called the major premise
and the second is known as the minor premise. Students were shown examples from high
school where they have drawn conclusions using universal modus ponens. The
application of the Pythagorean theorem is a very good illustration of this:
The Pythagorean theorem states that if you have any right-angled triangle with
hypotenuse c and legs a and b, then c 2 = a 2 + b 2 . If you are then given a particular
triangle in which the legs are 3 and 4 respectively then the hypotenuse c can be evaluated
as follows: c 2 = 3 2 + 42 = 25
∴c = 5

This was done to show students that logic is not an isolated topic, but is applied regularly
at all levels in mathematics. The use of universal modus ponens in proofs were illustrated
by means of a few examples.

Universal modus tollens forms an important ingredient of proof by contradiction. It has

the following valid argument form:
∀ x, if P( x) then Q ( x)
~ Q( x) for a particular
∴ ~ P(a)

The validity of universal modus tollens results from combining universal instantiation
with modus tollens. Exercises in formal and informal language were done to help
students to recognize the valid argument forms of modus ponens and modus tollens.
 57

 
5.4.4.3 Valid and invalid arguments of Quantified statements

An argument is called valid if, and only if, its  form is valid. To say that an argument form
is valid means that no matter what particular predicates are substituted for the predicate
symbols in its premises, if the resulting  premise statements are all true, then the
conclusion is also true. Specific examples of the converse and inverse error in quantified
 
form were used to make students aware of invalid argument forms. Diagrams were
utilized to test for the validity of argument forms. Fruitful discussions emanated from the
use of these diagrams. Students became quite adept at using these diagrams. Examples of
the use of these diagrams are the following:

Example 4
Determine whether the following arguments are valid or invalid. Support your answer by
drawing diagrams:

(a) All people are mice

All mice are mortal
∴ All people are mortal

(b) All healthy people eat an apple a day

Helen eats an apple a day
∴ Helen is a healthy person
(The diagrams are shown in figures 9 and 10)
 58

Solution  
(a)
 

 
Mice
  Mortals

Mice
People

people

Major premise
Minor premise

Mortals

Mice

people
People

Conclusion

Figure 9
 59

(b)  

People that eat an apple a   People that eat an apple a

day day

• Helen
healthy people

Major premise
Minor premise

People that eat an apple a People that eat an apple a

day day

• Helen

healthy people healthy people

• Helen

Conclusion A Conclusion B

∴ Argument is invalid since the conclusion does not necessarily follow from the
premises.
Figure 10
 60

Students got so adept at using these diagrams   that they changed the minor or major
premise to see if it will change the outcome. For example the following question where
diagrams were used to test the validity of the  argument:
No polynomial functions have horizontal asymptotes
This function has a horizontal asymptote  
∴ this function is not a polynomial
 
was changed to:
no polynomial function have horizontal asymptotes
this function does not have a horizontal asymptote
∴ this function is a polynomial function

Almost the whole class participated in the discussion that ensued. Some students used set
theory learnt in another class to solve the problem. They argued that there is no clear
conclusion for the changed question and that their diagrams showed this. The diagram is
shown below in figure 11.

Set of all functions

Polynomial Functions with horizontal

functions asymptotes

Figure 11

Based on student responses in class we were reasonably certain that they could cope with
questions dealing with quantified statements. We therefore gave them the following class
test on quantified statements:
 61

Rewrite the following in formal language:  

A (i) I trust every animal that belongs to me.
(ii) Dogs gnaw bones  
(iii) I admit no animals into my study unless they will beg when told to do so.
(iv)  
All the animals in the yard are mine.
(v) I admit every animal that I trust into my study.
 
(vi) The only animals that are really willing to beg when told to do so are dogs.
∴ All the animals in the yard gnaw bones.

B (i) When I work a logic example without grumbling, you may be sure it is one I
understand.
(ii) The arguments in these examples are not arranged in regular order like the
ones I am used to.
(iii) No easy examples make my head ache.
(iv) I can’t understand examples if the arguments are not arranged in regular
order like the ones I am used to.
(v) I never grumble at an example unless it gives me a headache.
∴ These examples are not easy.

Students did not perform very well in this test, but since the average was a pass mark we
decided to continue with similar exercises. In the subsequent lecture we gave the students
another test that again required them to rewrite sentences in formal language. We decided
that this test would double as a pre-test25, since we had a suspicion that all was not well
where these kinds of statements were concerned. The following is this test:

Rewrite the following sentences in formal language:

A (i) No birds, except ostriches, are nine feet high.

(ii) There are no birds in this aviary that belong to anyone but me.
(iii) No ostrich lives on mince pies.
(iv) I have no birds less than nine feet high.

B (i) All writers who understand human nature are clever.

(ii) No one is a true poet unless he can stir the hearts of men.
(iii) Shakespeare wrote Hamlet
(iv) No writer who does not understand human nature can stir the hearts of
men.
(v) None but a true poet could have written Hamlet.

To our amazement the students performed even worse in this test. The majority of
students had difficulty in rewriting sentences in formal language that contained the words
“no, none, unless, never and except” in it. We therefore decided to do intervention as the
understanding of quantified statements is crucial in understanding mathematical
statements. We gave them the following exercise which contained statements that
contained the words that students found problematic:
25
See Appendix C1
 62

 
Rewrite the following statements in formal language:
(i) No easy examples in mathematics are challenging.
(ii) No difficult problems in mathematics  can be solved easily.
(iii) No people except registered students are allowed to attend classes.
(iv)  
Some people will not go to church unless there is a special service.
(v) Some people will not buy clothes unless they have a birthday.
 
The solution to the above exercise was discussed and errors were pointed out and
corrected. Once we were satisfied that students could now cope with these types of
sentences we gave a follow-up test26 to determine if the erroneous reasoning was
corrected. The following is this test:
Rewrite the following statements in formal language:
(i) No bank closes before 3:30 unless it is a small bank.
(ii) No shark eats plankton unless it is a whale shark.
(iii) Students never study unless they have to prepare for a test.
(iv) None but a true gentleman will offer his seat to a lady on a bus.
(v) None but a brave soldier will fight in a war.

There was a dramatic improvement in student performance on this follow-up test hence
we were satisfied that students have now acquired the ability to deal with these kinds of
statements.

Students were now presented with a post-test for Puzzle I (the pre-test for this puzzle
was given in the very first lecture). The post-test is presented in Appendix A2. This
puzzle is equivalent to puzzle I as only the context was changed. If we compare the
puzzles we find the following similarities:
- Both puzzles had five statements
- The 1st , 2nd and 3rd statements of puzzle III contained if-then statements (compare
to puzzle 1 where statements 1, 2 and 5 was if-then statements)
- Statement 3 of puzzle III contained an or (compare to puzzle 1 where statement 4
contained the or)
- Both puzzles could be proven using modus ponens, modus tollens and then
disjunctive syllogism
- Both puzzles contained only compound statements.

We felt that students were also ready now to deal with forming of conclusions when
presented with arguments containing quantified statements. We therefore gave students
the following puzzle which served as a pre-test for arguments with quantified
statements. 27 The post-test for this puzzle was given approximately a month later.28
Again we made certain that the post-test was equivalent to the pre-test. Hence if we
compare the puzzles we note the following similarities:
- Both puzzles consist of 5 quantified statements and a given conclusion

26
See Appendix C2
27
See Appendix D1
28
See Appendix D2
 63

- In both cases students were required  to rearrange the statements so that the
conclusion follows logically
 
- In both cases 5 connections between statements needed to be made in order to arrive
at the conclusion
  proceeded to the section on proofs.
This concluded the section on logic. We then
 
5.5 The proof component of the course

Our intention with this section was to determine if there was a transfer of skills as far as
logic was concerned. In other words, to determine if the skills of making connections
between statements and forming conclusions from arguments would translate into better
deductive abilities and hence proving abilities of students. Stylianides et al [80] are of
the opinion that modus ponens and modus tollens form the basis of the methods of proof
that we utilized in our study as can be seen in the following quote: “Modus ponens is the
foundation of direct proof and the proof method by mathematical induction, whereas
modus tollens is the foundation of indirect proof (this includes the proof methods by
contradiction and by contraposition).”

What we therefore attempted to do in our study was for students to transfer the skills
learnt in the context of logic to the context of proving. Some researchers argue that this
kind of transfer is not possible. The experiment of reflex theorists Woodward and
Thorndike where adults who after special exercises could determine the length of short
lines, but could not transfer this skill to determine the length of long lines is a classic
example where transfer did not take place. A closer scrutiny of Woodward and
Thorndike however reveals that they do think that transfer is possible when the transfer
task and the learning task are identical. Bransford et al [7] argue that in this view of
learning transfer, where the emphasis is on identical elements of task, there is no
consideration of any learner characteristics such as where attention was directed,
whether relevant principals were extrapolated etc. they are of the opinion therefore that
in this view the primary emphasis is on drill and practice. Anderson et al [1] are of
similar opinion. Modern theories of learning and transfer maintained the emphasis on
practice, but they specify the kinds of practice that are important and take learner
characteristics like existing knowledge and strategies into account.

The approach we used to facilitate the transfer of skills from logic to proofs was by
means of a cue. Our motivation for doing so is based on the following quote from
Anderson et al [1]: “The amount of transfer depends on where attention is directed
during learning. Training on cues that signal the relevance of an available skill may
deserve much more emphasis than they now typically receive in instruction”. Cheng et
al [14] is of similar opinion as can be seen from the following quote: “Training was
effective only when abstract principles were coupled with examples of selection
problems, which served to elucidate the mapping between abstract principles and
concrete instances.” Our use of cues to signal the relevance of an available skill will be
demonstrated when the different methods of proof are discussed.
 64

5.5.1 Set theory  

 
Since the language of set theory is used in every mathematical subject and, in particular,
set identities play an important role in proving certain mathematical results, a brief
introduction of elementary set theory is given.  Certain set identities and set inclusions are
then provided to illustrate the use of rules of inference (given as cues) and to serve as an
 
introduction to proving in mathematics which will be dealt with in subsequent sections.

5.5.1.1 Definitions and Language of set theory

The words set and element are undefined terms of set theory just as sentence, true and
false are undefined terms of logic.
If S is a set then a ∈ S means that a is a member of S or an object in S.
a ∉ S means a is not a member of S or a does not belong to the set S.
We often write a set as follows: S = {x ∈ A / P( x)} where this means an element x is in
S if and only if x is in A and P(x) is true.
If A and B are two sets, A is called a subset of B, if and only if every element of A is also
an element of B. This is written as A ⊆ B
A is not a subset of the set B (written A ⊄ B ) if and only if, there is at least one element
of A that is not an element of B.
A is a proper subset of B, if and only if every element of A is in B, but there is at least
one element of B that is not in A.

5.5.1.2 Set Equality

Given sets A and B, A equals B (A = B), if and only if, every element of A is in B and
every element of B is in A. In symbols: A = B ⇔ A ⊆ B and B ⊆ A

5.5.1.3 Operations on sets

Let A and B be subsets of a universal set U, then the following hold:

(a) The union of A and B, denoted A ∪ B is the set of all elements x in U such that x
is in A or x is in B, symbolically: A ∪ B = {x ∈U / x ∈ A or x ∈ B} or
x ∈ A ∪ B ⇔ x ∈ A or x ∈ B
(b) The intersection of A and B denoted by A ∩ B , is the set of all elements x in U
such that x is in A and x is in B. Symbolically:
A ∩ B = {x∈U / x∈ A and x∈ B}

(c) The difference of B minus A (or the relative complement of A in B) denote B – A,

is the set of all elements x in U such that x is in B and x is not in A, symbolically:
B − A = {x ∈U / x ∈ B and x ∉ A}
 65

(d) The complement of A denoted A c, is the

  set of all elements x in U such that x is
not in A, symbolically: A = {x ∈U / x ∉ A}
c

 
5.5.1.4 Element method for proving that one
  set is a subset of another

Let sets X and Y be given. To prove that X ⊆ Y  , the following must be done:
(i) Suppose that x is a particular but arbitrarily chosen element of X,
(ii) Show that x is an element of Y

5.5.1.5 Element (basic) method for proving that sets are equal

Let sets X and Y be given. To prove that X = Y the following must be done:
(i) Prove that X ⊆ Y
(ii) Prove that Y ⊆ X

5.5.1.6 Proving by Division into Cases

One of the strategies that can be utilized to prove a mathematical statement is as

follows: Suppose we know that: A1 or A2 or A3 or … An is true. By definition of OR at
least one of the statements Ai is true (although we might not know which one). Suppose
you want to deduce a conclusion C. That is suppose you want to show that:
A1 or A2 or A3 or ... An → C Then prove all the implications:
A1 → C , A2 → C , A3 → C , ..., An → C and conclude that regardless of which statement Ai
happens to be true, the truth of C follows.

5.5.1.7 Empty set

The unique set with no elements is called the empty set. It is denoted by the symbol φ .
A set with no elements is a subset of every set. Symbolically: If φ is a set with no
elements and A is any set then φ ⊆ A .

Uniqueness of the empty set

There is only one set with no elements.

5.5.1.8 Disjoint sets

Two sets are called disjoint if, and only if, they have no elements in common.
Symbolically: A and B are disjoint ⇔ A ∩ B = φ .

Example 5
Let A = {1;3;5} and B = {2;4;6}
Then A ∩ B = φ
∴ A ∩ B is disjoint
 66

Example 6  
Show that A – B are B are disjoint sets.
 
Solution:
To show that A – B and B are two disjoint sets  we need to show that: ( A − B ) ∩ B = φ
Suppose x ∈ ( A − B) ∩ B , then x ∈ A − B and  x ∈ B - by definition of intersection
Therefore ( x ∈ A and x ∉ B) and x ∈ B - by definition of difference of two sets
Hence x ∈ A and ( x ∉ B and x ∈ B) - by associative law
So in particular x ∉ B and x ∈ B
This is a contradiction,
hence our supposition is false and hence ( A − B) ∩ B has no elements
therefore ( A − B) ∩ B = φ - uniqueness of empty set
it follows from the definition that A – B and B are disjoint

Mutually Disjoint Sets

The sets A1, A2, …,An are mutually disjoint ( or pairwise disjoint) if and only if, no two
sets Ai and Aj with i ≠ j have any elements in common.
Symbolically: For all i, j = 1, 2, ..., n Ai ∩ A j = φ if i ≠ j

Example 7
A1 = {3;5} , A2 = {1;4;6} , A3 = {2} , then
A1 ∩ A2 = φ , A1 ∩ A3 = φ , A2 ∩ A3 = φ
∴ A1 A2 A3 are mutually disjoint.

5.5.1.9 Partition of sets

A collection of non-empty sets {A1 , A2 , ..., An }is a partition of a set A if, and only if:
(i) A = A1 ∪ A2 ∪ ... ∪ An
(ii) A1, A2, …, An are mutually disjoint

Example 8
Let A = {1; 2 ; 3 ; 4; 5; 6 } , A1 = {1; 2} , A2 = {3; 4} and A3 = {5; 6}. Is {A1 , A2 , A 3 } a
partition of A?

Solution
(i) A = A1 ∪ A2 ∪ A3
(ii) A1 ∩ A2 = φ , A1 ∩ A3 = φ , A2 ∩ A3 = φ
∴ {A1 , A2 , A3 } is a partition of A.

5.5.1.10 Power Sets

Given a set A, the power set of A, denoted P(A), is the set of all subsets of A.
 67

Example 9  
Find the power set of the set {x; y} , that is, find P ({x; y})
 
Solution  
P ({x; y}) = {φ , {x}, {y}, {x; y}}
 
5.5.1.11 (a) Example 10 (Examples of proofs of set identities and set inclusions
where cues on logic are given)

Prove that for all sets A and B:

⎡make use of : p ∧ q ⎤
A∩ B ⊆ A ⎢
∴ p ⎥⎦
(i)
⎣

⎡make use of : p ⎤
A⊆ A∪ B ⎢
∴ p ∨ q ⎥⎦
(ii)
⎣

⎡make use of the rule of inf erence : p ∨ q ⎤

(iii) A − B = A − ( A ∩ B) ⎢⎢ ~ p ⎥⎥
⎢⎣ ∴q ⎥⎦

⎡make use of : p ⎤
A ∪ ( B ∩ C ) = ( A ∪ B) ∩ ( A ∪ C ) ⎢
∴ p ∨ q ⎥⎦
(iv)
⎣

Solutions
(i) Suppose x ∈ A ∩ B . Then x∈ A and x∈ B and so in particular x∈ A . This
shows that A ∩ B ⊆ A .

(ii) Suppose x∈ A . Then x∈ A or x∈ B and so x ∈ A ∪ B . Hence A ⊆ A ∪ B .

(iii) Suppose x ∈ A − ( A ∩ B) . Then x∈ A and x ∉ A ∩ B

Hence x∈ A and (either x ∉ A or x ∉ B)
Therefore x∈ A and x∉ B
Hence x ∈ A − B ; proving that A − ( A ∩ B) ⊆ A − B
Conversely suppose x ∈ A − B , then x∈ A and x∉ B
⎡ p ⎤
Hence x ∈ A and x ∉ A ∩ B ⎢ ⎥
⎣∴ p ∨ q⎦
and so x ∈ A − ( A ∩ B)
∴ A − B ⊆ A − ( A ∩ B)
∴ A − B = A − ( A ∩ B)
 68

(iv) Suppose x ∈ A ∪ ( B ∩ C ) . Then x∈  A or x ∈ B ∩ C

Case 1: x∈ A
 
⎡p ⎤
Since x ∈ A , x ∈ A ∪ B and x ∈ A ∪ C ⎢ ⎥
  ⎣∴ p ∨ q ⎦
Hence x ∈( A ∪ B ) ∩ ( A ∪ C ) and so A ∪ ( B ∩ C ) ⊆ ( A ∪ B) ∩ ( A ∪ C )
 
Case 2: x ∈ B ∩ C
Since x ∈ B ∩ C , x ∈ B and x ∈ C
⎡p ⎤
Hence x ∈ A ∪ B and x ∈ A ∪ C ⎢ ⎥
⎣∴ p ∨ q ⎦
Therefore x ∈( A ∪ B) ∩ ( A ∪ C )
And so A ∪ ( B ∩ C ) ⊆ ( A ∪ B) ∩ ( A ∪ C )
Therefore regardless of which case is true, A ∪ ( B ∩ C ) ⊆ ( A ∪ B) ∩ ( A ∪ C )

Conversely: suppose x ∈( A ∪ B) ∩ ( A ∪ C )
Then x ∈ A ∪ B and x ∈ A ∪ C and so
( x∈ A or x∈ B ) and ( x∈ A or x∈ C )
Case 1: x∈ A
⎡p ⎤
If x ∈ A , then x ∈ A ∪ ( B ∩ C ) ⎢ ⎥ , so
⎣∴ p ∨ q ⎦
( A ∪ B) ∩ ( A ∪ C ) ⊆ A ∪ ( B ∩ C )
Case 2: x∉ A
Then x ∈ B and x ∈ C , so x ∈ B ∩ C
⎡p ⎤
And hence x ∈ A ∪ ( B ∩ C ) ⎢ ⎥
⎣∴ p ∨ q ⎦
∴ ( A ∪ B) ∩ ( A ∪ C ) ⊆ A ∪ ( B ∩ C )
∴ A ∪ ( B ∩ C ) = ( A ∪ B) ∩ ( A ∪ C )

(b) The following list of set identities can be proved in the same way as the above
examples.
Let all sets referred to below be subsets of a universal set U.

(i) Commutative laws: for all sets A and B

(a) A ∩ B = B ∩ A and (b) A∪ B = B ∪ A

(ii) Associative laws: for all sets A, B and C

(a) ( A ∩ B) ∩ C = A ∩ ( B ∩ C ) and (b) ( A ∪ B) ∪ C = A ∪ ( B ∪ C )

(iii) Distributive laws: For all sets A, B and C

(a) A ∪ ( B ∩ C ) = ( A ∪ B) ∩ ( A ∪ C ) and
(b) A ∩ ( B ∪C ) = ( A ∩ B) ∪ ( A ∩ C )
 69

(iv) Intersection with U for all sets A:  

A ∩U = A
 
(v) Double Complement Law: For all  sets A
(A c ) c = A
 
(vi) Idempotent laws: For all sets A
(a) A ∩ A = A and (b) A∪ A = A

(vii) De Morgan’s Laws: For all sets A and B

(a) ( A ∪ B) c = A c ∩ B c and (b) ( A ∩ B) c = A c ∪ B c

(viii) Union with U: A ∪U = U

(ix) Absorption Laws: For all sets A and B

(a) A ∪ ( A ∩ B) = A and (b) A ∩ ( A ∪ B) = A

(x) Alternate representation for set difference: For all sets A and B
A− B = A ∩ B c

Proving set identities using the list in 5.5.1.11 (b)

For example the statement “For all sets A, B and C, ( A − B) ∩ (C − B) = ( A ∩ C ) − B ” is

proved by applying the set identities in a step by step fashion and giving a reason for each
step as can be seen below.

( A − B) ∩ (C − B)
= ( A ∩ B c ) ∩ (C ∩ B c ) , alternate representation for set difference.
= ( A ∩ B c ) ∩ ( B c ∩ C ) , commutative law.
= ( A ∩ B c ) ∩ B c ) ∩ C , associative law.
= ( A ∩ ( B c ∩ B c )) ∩ C , associative law.
= ( A ∩ B c ) ∩ C , idempotent law.
= A ∩ ( B c ∩ C ) , associative law.
= A ∩ (C ∩ B c ) , commutative law.
= ( A ∩ C ) ∩ B c , associative law.
= ( A ∩ C ) − B , alternate representation for set difference.
Students were given a number of exercises to consolidate the theory on sets. Once we
were certain that they have acquired the necessary skills we gave them a pre-test29 on
proofs in set theory. Approximately a week later they were given an equivalent

29
See Appendix E1
 70

post-test30. Students also wrote a class test on  proofs of set theory, this test formed part of
their course mark but was not utilized as a measuring instrument in our study. The June
examinations also contained questions where  students were required to prove statements
on sets using both the element argument and set identities.
 
5.5.2 Elementary Number Theory
 
Before continuing the discussion on different methods of proof, we give a brief outline of
some of the elementary properties of numbers. This is because the examples that will be
used to explain the proof methods involve number theoretic properties. We first give
some definitions and examples of the different concepts.

Definitions

An integer n is even if, and only if, n = 2k for some integer k.

An integer n is odd if, and only if, n = 2k + 1 for some integer k

An integer n is prime if, and only if n >1 and for all positive integers r and s, if n = r s
then r = 1 or s = 1

An integer n is composite if, and only if n = r × s for some positive integers r and s with
r ≠ 1 and s ≠ 1
a
A real number r is rational if, and only if r = for some integer a and b with b ≠ 0
b
A real number is irrational if it is not rational.

Floor
Given any real number x, the floor of x denoted ⎣x ⎦ , is defined as follows:
⎣x ⎦ = that unique integer n such that n ≤ x < n + 1 .

Ceiling
Given any real number x, the ceiling of x denoted ⎡x ⎤ , is defined as follows:
⎡x ⎤ = that unique integer n such that n −1 < x ≤ n

Composite number
If n is an integer that is greater than 1, then n is composite if and only if ∃ positive
integers r and s such that n = r.s and r ≠ 1 and s ≠ 1

Composite number lemma

Let n be an integer that is greater than 1. If n is composite then n = r.s where
1 < r < n and 1 < s < n

30
See Appendix E2
 71

Common divisor  
Let a and b be two non-zero integers. Then there is a unique integer d such that:
(i) d is a common divisor of a and  b, that is d/a and d/b and
(ii) if c is an integer which is also a common divisor, that is c/a and c/b then
c/d.  
We call d the greatest common divisor of a and b.
 
Example 11
Prove that 0 is an even number

Solution
0 = 2. 0
∴0 is even by definition of even

Example 12
Prove that – 301 is odd

Solution
-301 = 2 x (-151) + 1
∴− 301 is odd by definition of odd

Example 13
Is 1 a prime number?

Solution
1 is not a prime number, since any prime number is greater than 1

Example 14
Is it true that every integer greater than 1 is either prime or composite?

Solution
Yes, since the two definitions are negations of one another.

If n and d are integers and d ≠ 0 , then n is divisible by d if, and only if, n = dk for
some integer k. The notation d/n is read “d divides n”

Quotient – Remainder theorem

Given any integer n and positive integer d, there exist unique integers q and r such that
n = d ⋅ q + r and 0 ≤ r < d
Example 15
n = 17; d = 3, then
17 = 3 × 5 + 2

d q r
 72

5.5.3 Methods of Proof  

 
Each method of proof follows a fixed procedure. Unlike the section on logic where we
gave the students a pre-test before teaching them, we first taught the students what the
procedure was for a specific method of proof  and then gave the pre-test. Using the errors
in reasoning of the pre-test as a guide we proceeded to teach the students again so as to
 
eliminate these errors. Once we were satisfied that they were now competent with a
specific method of proof we administered the post-test. The reason for this was because
students had no previous experience of these methods of proof and as previously
discussed, the students need cognitive structures that already exist for the process of
assimilation and accommodation to occur.

We now proceed to illustrate the different methods of proof by examples.

5.5.3.1 The method of direct proof

The method of direct proof is based on the method of generalizing from the generic
particular. In a direct proof you start with the hypothesis of a statement and make one
deduction after another until you reach the conclusion. This implies the following:

To prove a statement of the form ∀ x∈ D, if P( x) then Q( x) you suppose x is a

particular but arbitrarily chosen element of D that satisfies P(x) and then you show that
x satisfies Q(x).

The steps involved in the method of direct proof can therefore be stated as follows:

(i) Express the statement to be proved in the form ∀ x∈ D, if P( x) then Q( x)

(ii) Start the proof by supposing x is a particular but arbitrarily chosen element of D
for which the hypothesis P(x) is true
(iii) Show that the conclusion Q(x) is true by using definitions, previously established
results and the rules of inference

It is clear that in the case of direct proof the first two steps are easy to learn. The third
step however requires the student to be creative in order to prove the conclusion. This
third step therefore was our focus in the teaching process where we again emphasized
step – by- step reasoning as a cognitive tool.

Example 16
Prove that the product of any two odd integers is odd

Solution
Step 1
∀ m, n ∈ Ζ , if m and n are odd then m. n is odd
Step 2
Suppose m and n are particular but arbitrarily chosen integers so that m and n are both
odd.
 73

Step 3  
Then by definition of odd n = 2k + 1 and m = 2r + 1 for some integers k and r
  and so m. n = 4kr + 2k + 2r + 1
Then m. n = (2k + 1)(2r + 1) , by substitution
= 2(2kr + k + r) + 1, by applying the rules of algebra. Let s = 2kr + k + r.
 
Now s is an integer because products and sums of integers are integers and 2, r and k are
all integers. Hence m.n = 2 s + 1 and so by definition of odd, m.n is odd.
 
Example 17
Prove that the sum of any two rational numbers is rational.

Solution
Step 1
∀ real numbers r and s, if r and s are rational then r + s is rational

Step 2
Suppose r and s are particular but arbitrarily chosen rational numbers.

Step 3
a c
Then r = and s = for some integers a, b, c and d with b ≠ 0 and d ≠ 0 , by
b d
definition of rational. Therefore

a c ad + bc
r+s= + =
b d bd

Then ad + bc and bd are integers because products and sums of integers are integers and
because a, b , c and d are all integers. Also bd ≠ 0 , by the zero product property.
Thus r + s is rational by definition of rational numbers.

Example 18
Prove that for all integers a, b and c if a b and b/c then a/c

Solution
Step 1
∀ a, b, c ∈ Ζ , if a/b and b/c, then a/c

Step 2
Suppose a, b and c are particular but arbitrarily chosen integers such that a/b and b/c

Step 3
Since a/b, b = ak for some integer k and since b/c, c = bs for some integer s.
By substitution c = (a.k).s = a(k.s) , by the associative law for multiplication.
Let p = k.s. Then p is an integer since it is a product of integers and therefore c = a.p
where p is an integer and thus a divides c ,by definition of divisibility.
 74

Example 19 (method of division into cases) 

Prove that any two consecutive integers have opposite parity.
 
Solution
Two integers are called consecutive if, and  only if one is one more than the other.
Opposite parity means one integer is odd and the other is even.
 
Suppose that m and m + 1 are two particular but arbitrarily chosen consecutive integers.
By the parity property, either m is even or m is odd.

Case 1: (m is even)

In this case m = 2k for some integer k, and so m + 1 = 2k + 1, which is odd. Hence in

this case one of m and m + 1 are even and the other is odd.

Case 2: (m is odd)

In this case m = 2k + 1 for some integer k, and so

m + 1 = (2k + 1) +1 = 2k + 2 = 2(k + 1), but k + 1 is an integer because it is a sum of
two integers. Therefore m + 1 equals twice some integer and thus m + 1 is even. Hence
in this case also one of m and m + 1 is even and the other is odd.
It follows that regardless of which case actually occurs for the particular m and m + 1 that
are chosen, one of m and m + 1 is even and the other is odd

Example 20 (Application of Quotient-Remainder theorem)

Prove that the square of any odd integer has the form 8m + 1 for some integer m.

Solution
Step 1 (formal restatement)
∀ odd integers n, ∃ an integer m such that n 2 = 8m + 1

Step 2
Suppose n is a particular but arbitrarily chosen odd integer
Step 3
Then n can be written in one of the forms: 4q or 4q + 1 or 4q + 2 or 4q + 3 for some
integer q.
Now since n is odd and 4q and 4q + 2 are even n must have one of the forms: 4q + 1 or
4q + 3

Case 1: (n = 4q + 1 for some integer q)

Since n = 4q + 1
n 2 = (4q + 1)2
n 2 = 16q 2 + 8q + 1
n2 = 8(2q 2 + q) + 1
let m = 2q 2 + q, then m is an integer since 2 and q are integers and sums and products are
integers, so substituting:
 75

n 2 = 8m + 1 where m is an integer  

Case 2: (n = 4q + 3 for some integer q)  

Since n = 4q + 3
n 2 = (4q + 3) 2  
2 2
n = 16q + 24q + 9
 
n 2 = 16q 2 + 24q + (8 + 1)
n 2 = 8(2q 2 + 3q + 1) + 1
let m = 2q 2 + 3q + 1, then m is an integer since 2, 3 and q are integers and sums and
products of integers are integers. Hence substituting:
n 2 = 8m + 1 where m is an integer.

Case 1 and 2 show that given any odd integer, whether of the form 4q + 1 or 4q + 3,
n 2 = 8m + 1, which is what we needed to show.

Example 21 (Floor)
Prove that for all real numbers x and for all integers m, ⎣x + m⎦ = ⎣x ⎦ + m

Solution
Step 1
∀ x, m ∈ ℜ if x ∈ℜ and m∈ Ζ, then ⎣x + m ⎦ = ⎣ x ⎦ + m

Step 2
Suppose x is a particular but arbitrarily chosen real number and m is a particular but
arbitrarily chosen integer. Let n = ⎣x ⎦

Step 3
Then n is an integer and n ≤ x < n + 1 - by definition of floor
Add m to all sides to obtain: n + m ≤ x + m < n + m + 1
Now n + m is an integer since n and m are integers and a sum of integers is an integer,
and so by definition of floor: ⎣x + m ⎦ = n + m , but n = ⎣x ⎦ , hence by substitution
⎣x + m⎦ = ⎣x ⎦ + m , which is what was to be shown.

Example 22 (Floor)
⎧n
if n is even
⎢ n ⎥ ⎪⎪
Prove that for any integer n, ⎢ ⎥ = ⎨ 2
⎣ 2 ⎦ ⎪ n − 1 if n is odd
⎪⎩ 2
 76

Solution  
Suppose n is a particular but arbitrarily chosen integer. By the quotient-remainder
theorem, n is odd or n is even that is: n = 2q  or n = 2q + 1 for some integer q.

n  
Case 1: n = 2q ⇒ q =
2  
Then
⎢ n ⎥ ⎢ 2q ⎥
⎢ 2 ⎥ = ⎢⎣ 2 ⎥⎦
⎣ ⎦
= ⎣q ⎦
=q
n
=
2
As was to be shown.

n −1
Case 2: n = 2q + 1 ⇒ =q
2
Now
⎢ n ⎥ ⎢ 2q + 1⎥
⎢⎣ 2 ⎥⎦ = ⎢ 2 ⎥
⎣ ⎦
⎢ 1⎥
= ⎢q + ⎥
⎣ 2⎦
= q
n −1
=
2
This is what was required.

Example 23 (Ceiling)
⎡n 2 ⎤ n 2 + 3
Prove that for any odd integer n, ⎢ ⎥ =
⎢ 4 ⎥ 4

Solution
Step 1
⎡n 2 ⎤ n 2 + 3
∀ odd integers n, ⎢ ⎥ =
⎢ 4 ⎥ 4
Step 2
Suppose n is a particular but arbitrarily chosen odd integer. Then n = 2k + 1 for some
integer k.
 77

Step 3  
Hence
⎡ n 2 ⎤ (2k + 1) 2  
⎢ ⎥=
⎢ 4 ⎥ 4  
⎡ 4k + 4k + 1⎤
2
=⎢ ⎥  
⎢ 4 ⎥
⎡ 1⎤
= ⎢k 2 + k + ⎥
⎢ 4⎥
⎡n 2 ⎤
∴⎢ ⎥ = k 2
+ k +1
⎢ 4 ⎥

also
n 2 + 3 (2k + 1) 2 + 3
=
4 4
n + 3 4k + 4k + 1 + 3
2 2
=
4 4
n + 3 4k + 4k + 4
2 2
=
4 4
n +3
2
= k 2 + k +1
4

∴ LHS = RHS, this is what was to be shown.

Students were provided with a number of exercises where they had to prove statements
using the direct method of proof.

5.5.3.2 Disproof by counterexample

To disprove a statement of the form ∀ x ∈ D, if P( x) then Q( x) find a value of x in D

for which P(x) is true and Q(x) is false. Such an x is called a counterexample.

Example 24
Disprove the following statement by finding a counterexample:
∀ real numbers a and b , if a2 = b2 then a = b

Solution
Let a = 1 and b = -1
Then a2 = 12 = 1 and b2 = (-1)2 = 1 and so a2 = b2, but a ≠ b since 1 ≠ − 1

Students were required to practice and hence consolidate what was learnt by doing
similar exercises.
 78

5.5.3.3 Method of contradiction  

Epp [33] states that argument by contradiction   is based on the fact that either a statement
is true or false but not both. Thus the point of departure for a proof by contradiction is the
supposition that the statement to be proved   is false and the goal is to reason to a
contradiction. Hence as previously indicated the method of proof by contradiction
 
consists of the following steps:
(i) Suppose the statement to be proved is false (i.e. use negation)
(ii) Show that this supposition leads logically to a contradiction
(iii) Conclude that the statement to be proved is true.

In each method of proof there is a part where the student has to apply prior knowledge of
various types in order to do the proof and this is what we attempted to improve in
students. In other words we tried to get them to the point where they became creative in
their use of prior knowledge and hence to apply knowledge that we had not taught them.
The following two examples will illustrate this.

Example 25
Prove that the sum of any rational number and any irrational number is irrational

Solution
Step 1: Suppose not [here the student is supposed to take the negation of the statement
and suppose it to be true]

Suppose there is a rational number r and an irrational number s such that r + s is

rational.

Step 2: We must now induce a contradiction

a c
By definition of rational r = and r + s = for some integers a, b , c and d with
b d
b ≠ 0 and d ≠ 0 . By substitution we have:
a c
+ s = and so
b d
c a
s= −
d b
bc − ad
=
bd
Now bc – ad and bd are both integers and bd ≠ 0
Hence s is a quotient of the two integers bc – ad and with bd ≠ 0
So by definition of rational s is rational
This contradicts the supposition that s is irrational.

Step 3:
Hence the supposition is false and the statement is true.
 79

Example 26  
Prove that there is no greatest integer.
 
Solution
Step 1:  
Suppose there is a greatest integer N
 
Step 2:
Since N is the greatest integer N ≥ n for every integer n
Let M = N + 1
Now M is an integer since it is a sum of integers
Also M > N since M = N + 1
Thus M is an integer that is greater than the greatest integer, which is a contradiction.

Step 3:
This contradiction shows that the supposition is false and hence the statement is true.

Now if we compare the two examples we notice that step 1 and 3 are similar in both
cases. In step 1 in both cases the negation of the given statement was assumed to be true.
Step 3 in both cases stated that since a contradiction was induced the supposition was
false and hence the original statement was true. It is therefore easier to teach step 1 and 3
since it follows the same pattern every time.

Step 2 requires the student to use some kind of prior knowledge. If we compare step 2 of
the two examples we see that a different technique was utilized to induce a contradiction
in each case (in other methods of proof it might not be step 2, but a similar argument
holds). This is what we wanted to improve in our students i.e. the ability to be creative in
their use of prior knowledge. We also wanted to inculcate in them the notion of one-step-
reasoning. Our argument therefore is that we cannot teach the students all the different
techniques that are necessary to do step 2 since there is such a vast number of
mathematical statements. The best we can do is to teach them how to go about finding the
specific strategy to be employed in order to find that specific technique. In other words to
teach them the cognitive processes that is needed to get to the solution for the specific
mathematical proof. Epp [33] explains this as follows: “In order to evaluate the truth or
falsity of a statement one needs to understand what the statement is about. You need to
know the meanings of all terms that occur in a statement since mathematicians define
terms carefully and precisely.”

Example 27
Prove that for all integers n, if n 2 is even then n is even.

Solution
Step 1
Suppose there exist an integer n such that n 2 is even and n is odd.
 80

Step 2  
2
Suppose n is an integer such that n is even but n is odd.
Now n odd implies n = 2k + 1 for some integer  k, by definition.
[Teaching strategy - What is the next step? Look for cues in the problem. Observe that n2
appears, so calculate it]  
Hence n = (2k + 1) = 4k + 4k +1 = 2(2k2 + 2k) + 1 = 2l + 1,where l is the integer
2 2 2
 
2k2 + 2k . But then n2 is odd.
So we have n2 is even and n2 is odd, a contradiction.

Step 3
Hence the supposition is false and the statement is true.

Example 28
Prove that 2 is irrational

Solution
Step 1
Suppose 2 is rational

Step 2
Then there are integers m and n with no common factors so that:
m
2= …(1)
n
Squaring both sides gives:
m2
2= 2 …(2)
n
Or equivalently: m 2 = 2n 2 …(3)

Then m 2 is even, by definition of even. It follows that m is even by example 27.

Then m = 2k for some integer k …(4)

Substituting (4) into (3) gives:
m 2 = (2k) 2 = 4k 2 = 2n 2 …(5)
dividing both sides of the right –most equation of (5) gives:
n 2 = 2k 2
2
Consequently n is even and so n is even by example 27.
But we also know that m is even.
Hence both m and n have a common factor of 2, but this contradicts the supposition that
m and n have no common factors.

Step 3
Hence the supposition is false and so the statement is true.
 81

5.5.3.4 Method of proof by contraposition 

 
This method is based on the logical equivalence between a statement and it’s contra-
positive. To prove a statement by contraposition, you take the contra-positive of the
 
statement, prove the contra-positive by a direct proof and conclude that the original
statement is true. The method of proof by contraposition therefore consists of the
 
following steps:
(i) Express the statement to be proved in the form: ∀ x∈ D, if P( x) then Q( x )
(ii) Rewrite this statement in the contra-positive form:
∀ x∈ D, if ~ Q( x) then ~ P( x)
(iii) Prove the contra-positive by a direct proof:
(a) Suppose x is a particular but arbitrarily chosen element of D such that Q(x)
is false.
(b) Show that P(x) is false.

Example 29
Prove that the negative of any irrational number is irrational.

Solution
Step 1
∀ real numbers x, if x is irrational then –x is irrational.

Step 2
∀ real numbers x, if –x is not irrational then x is not irrational OR equivalently:
Since –(-x) = x: ∀ real numbers x, if x is rational then –x is rational.

Step 3
(a) Suppose x is any particular but arbitrarily chosen rational number.
a
(b) By definition of rational x = for some integers a and b with b ≠ 0
b
⎛ a ⎞ −a
Then − x = − ⎜ ⎟ =
⎝b⎠ b
Since both –a and b are integers and b ≠ 0 , -x is rational, by definition of rational,
as was to be shown.

Example 30
It was mentioned earlier that the use of cues was employed as a strategy throughout our
teaching on methods of proof. We show how logical equivalence of statement forms can
be used in a proof by contra-position. Consider the following problem.
Use proof by contra-position to show that the difference of any rational number and any
irrational number is irrational. Make use of the following logical equivalence in your
proof by contraposition: p → q ∨ s ≡ p ∧ ~ q → s
Solution
Here the cue was given by the logical equivalence and the idea was to get students to re-
write the statement using the equivalence learnt in logic, in such a way so as to make it
 82

 
possible to prove the statement by direct proof which would otherwise be very difficult
to do.
 
The formal statement is given by:
  s is irrational , then r − s is irrational
∀ real numbers r and s, if r is rational and
 
The contra-position of this statement is:
∀ real numbers r and s, if r − s is rational then r is irrational or s rational

Using the logical equivalence the statement becomes:

∀ real numbers r and s, if r − s is rational and r is rational , then s is rational
Suppose r and s are particular but arbitrarily chosen real numbers such that r-s and r is
rational
a c
Then r = and r − s = for some integers a,b,c and d
b d
with b ≠ 0 and d ≠ 0 - by definition of rational,
then
c
− s = −r
d
c
s =r−
d
a c
s= −
b d
ad − bc
s=
bd

But ad – bc are integers because a,b,c and d are integers and products and differences of
integers are integers. Also bd ≠ 0 sin ce b ≠ 0 and d ≠ 0
Thus by definition of rational s is rational as was required to prove.

5.5.3.5 Connection between proof by contradiction and proof by contraposition

Epp [33] argues about the connection between proof by contradiction and proof by
contraposition as follows:
“In a proof by contraposition the statement ∀ x∈ D, if P ( x) then Q( x) is proved by
giving a direct proof of the equivalent statement: ∀ x∈ D, if ~ Q( x) then ~ P( x ) .
To do this, you suppose that you are given an arbitrary element of x in D such that
~ Q(x). You then show ~ P(x).
Exactly the same sequence of steps can be used as the heart of a proof by contradiction
for the given statement. The only thing that changes is the context in which the steps are
written down. To rewrite the proof as a proof by contradiction you suppose there is an x
in D such that P(x) and ~ Q(x). You then follow the steps of the proof by contraposition
to deduce the statement ~ P(x). But ~P(x) is a contradiction to the supposition that P(x)
and ~ Q(x).”
 83

  with proof by contradiction you can “cut

The reverse of this is also true i.e. if you start
and paste” a portion of this proof to form the heart of proof by contraposition. We have
 
employed this strategy i.e. start with contradiction and then cut and paste for
contraposition. The following example illustrates this:
 
Example 31
 
Prove the following statement first by contradiction and then by contraposition:
For all integers a, b and c, if a ∤ bc then a ∤ b

Solution
Proof by contradiction
Step 1
Suppose ∃ integers a, b and c such that a ∤ bc and a/b

Step 2
Since a/b there exists an integer k such that b = ak ,by definition of divide
Then bc = (ak ).c
bc = a (k .c) , by associative law
But kc is an integer since it is a product of integers and so a/bc - by definition of divide
Thus a ∤ bc and a/bc which is a contradiction

Step 3
Hence the supposition is false and the statement is true.

Proof by contraposition
Contraposition of statement: For all integers a, b and c if a/b then a/bc
Suppose a, b and c are particular but arbitrarily chosen integers such that a/b,
then b = ak - by definition of divide
then bc = (ak ).c
bc = a (k .c) - by associative law
But kc is an integer since it is a product of integers and so a/bc - by definition of divide
,as was to be shown.
The above example shows how parts of the proof by contradiction can be “cut and paste”
to form the heart of proof by contraposition. [The part in the border represents the part
that was cut and pasted]
In order to consolidate the above method of proof exercises were given that contained
similar examples. Students completed pre-tests and post-tests on direct proof based on
divisibility31 and number theory32 respectively. Although we also did pre- and post-tests
for the proof by contradiction and contraposition we could not include it in our study
because of an error in the statement of the problem. The students also wrote a class test
on direct and indirect proof. Upon the completion of the section on direct and indirect
proof we proceeded to the principal of mathematical induction.

31
See Appendices F1 and F2.
32
See Appendices G1 and G2.
 84

 
5.5.4 Mathematical Induction
 
Mathematical induction is used to check conjectures about the outcomes of processes that
occur repeatedly and according to definite   patterns. The proof of a statement by
mathematical induction consists of two steps:
 
(i) In step 1 (basis step) you prove that P(a) is true for a particular integer a.
(ii) In step 2 (inductive step) you prove that for all integers k ≥ a , if P(k) is true
then P(k + 1) is true. To prove step 2 we do the following:
(a) Suppose that P(k) is true where k is a particular but arbitrarily chosen
integer ≥ a . This supposition that P(k) is true , is called the inductive
hypothesis.
(b) Then we prove that P(k + 1) is true using the inductive hypothesis. Finally
we can conclude that P(n) is true for all integers n ≥ a . The principal of
mathematical induction can be formally represented as follows:

Let P (n) be a predicate that is defined for integers n, and let a be a fixed integer.
Suppose the following two statements are true:
(i) P(a) is true
(ii) For all integers k ≥ a , if P (k) is true then P (k + 1) is true.
Then the statement for all integers n ≥ a , P(n) is true.

Example 32
n (n + 1)
Use mathematical induction to prove that: 1 + 2 + ... n = for all integers n ≥ 1
2
Solution
n (n + 1)
To start one must first identify P (n). In this case P (n): 1 + 2 + ... n =
2
Basis step [We must show P(1) is true]
LHS: P(1) = 1
1(1 + 1) 2
RHS: P (1) = = =1
2 2
∴ LHS = RHS
And so the formula is true for n = 1.

Inductive step
(a) Suppose that k is a particular but arbitrarily chosen integer greater than or equal
k (k + 1)
to 1 such that P(k) is true: P(k ) :1 + 2 + ... + k =
2
(b) Next we must prove P (k+1) is true, that is:
(k + 1)(k + 2)
1 + 2 + 3 + ... + k + (k + 1) =
2
Now P (k+1) = 1 + 2 + 3 + … + k + (k + 1)
 85

k (k + 1)  
= + k +1
2
 
k +1
= (k + 2)
2  
k +1
= [(k + 1) + 1]  
2
This proves that P (k + 1) is true.
n (n + 1)
∴ We conclude therefore that 1 + 2 + ... n =
2
Example 33
Use mathematical induction to prove the following statement:
2
⎡n ⎤
1 + 2 + ... + n = ⎢ (n + 1)⎥
3 3 3

⎣2 ⎦

Solution
Basis step
P(1): LHS = 1 3 = 1
⎡1(1 + 1) ⎤
2 2
⎡2⎤
RHS = ⎢ ⎥ = ⎢ ⎥ = (1) 2 = 1
⎣ 2 ⎦ ⎣2⎦
∴LHS = RHS
∴formula is true for n = 1

Inductive step
(a) Suppose k is a particular but arbitrarily chosen integer greater than or equal to 1,
2
⎡k ⎤
such that P (k) is true: P (k ) :1 + 2 + ... + k = ⎢ (k + 1)⎥
3 3 3

⎣2 ⎦
(b) Now we need to prove it is true for P(k+1):
P (k ) :13 + 2 3 + ... + k 3 + (k + 1) 3
2
⎡k ⎤
= ⎢ (k + 1)⎥ + (k + 1)
3

⎣2 ⎦
(k + 1)
[ ]
2
= k 2 + 4(k + 1)
4

=
(k + 1)
2
[
k 2 + 4k + 4 ]
4

=
(k + 1)
2
[
(k + 2)2 ]
4
⎡ (k + 1)(k + 2) ⎤
2

= ⎢ ⎥⎦
⎣ 2
This prove that P (k+1) is true.
 86

  2
⎡n ⎤
We conclude therefore that: 1 3 + 2 3 + ... + n 3 = ⎢ (n + 1)⎥ ∀ n ≥1
  ⎣2 ⎦

 
5.5.4.1 Using mathematical induction to prove a divisibility property

Example 34  
Prove by mathematical induction that for all integers n ≥1, 2 2 n − 1 is divisible by 3.

Solution
Basis step
P (1): 2 2 (1) − 1 = 4 − 1 = 3 , which is divisible by 3.
∴ P (1) is true.

Inductive step
(a) Inductive hypothesis
Suppose that for any particular but arbitrarily chosen integer ≥ 1 , P (k) is true
i.e. 2 2 k − 1 is divisible by 3.
(b) We now need to show that P(k+1) is true. That is we need to show that
2 2 ( k +1) − 1 i.e. 2 2 k + 2 − 1 is divisible by 3.
2 2 ( k +1) − 1 = 2 2 k + 2 − 1
= 2 2 k .2 2 − 1
Now
= 4.2 2 k − 1
= 4.(2 2 k − 1) + 3
Now since 2 2 k − 1 is divisible by 3 by the inductive hypothesis let 2 2 k − 1 = 3l for some
integer l ,then
2 2 ( k +1) = 4 ⋅ 3l + 3
= 3(4l + 1)
∴ 2 2 ( k +1) − 1 is divisible by 3
∴ P (k+1) is true.
Therefore 2 2 n − 1 is divisible by 3 ∀ n ≥ 1

5.5.4.2 Using mathematical induction to prove an inequality

Example 35
Use mathematical induction to prove that n 2 < 2 n ∀ n ≥ 5

Solution
Basis step
P (5): LHS = 52 = 25
RHS = 25 = 32. And so 25 < 32 ∴ P (5) is true
Inductive step
 87

(a) Suppose that k is a particular but  arbitrarily chosen integer ≥ 5 such that P (k)
is true. That is k 2 < 2 k .
 
We now need to prove that P (k+1) is true. That is (k + 1) < 2 k + 1
2
(b)
Now  
(k + 1) = k + 2k + 1 < 2 + 2k  + 1
2 2 k

(k + 1)2 < 2 k + 2 k = 2.2 k = 2 k +1

∴ (k + 1) < 2 k +1
2

∴P(n) is true ∀ n ≥ 5

The section on ordinary mathematical induction was concluded with vigorous discussion
on the strategies involved in proving by mathematical induction. To consolidate exercises
were given and immediate feedback on the exercises was provided. A pre- and post-test
on the method of ordinary induction was administered.33

5.5.4.3 Strong Mathematical Induction

The lesson was started with a definition of recursively defined sequences, since at this
point students did not yet deal with recursive sequences.

Definition
A sequence a 0 , a1 , a 2 , ... is said to be given recursively if the first few terms are specified
and a rule (called a recursion) is given for computing each later term from the earlier
ones.

5.5.4.4 Teaching to connect ordinary mathematical induction to strong induction

Instead of starting the lecture in the normal way by discussing the steps involved in
proving by strong mathematical induction, the lecturer used an innovative teaching
strategy to link ordinary mathematical induction with strong mathematical induction.
This strategy is described in the next example. The lecturer started the example and
stopped where it is required to prove that the general formula represents the recursive
formula.

Example 36
Suppose that a 0 , a1 , a 2 , ... is the sequence such that a 0 = 1, a1 = 2 and
a n + 2 = 2a n + a n +1 ∀ n ≥ 0
Then a 0 = 1, a1 = 2 and
a 2 = 2a0 + a1 = 2(1) + 2 = 4 = 2 2

33
See Appendices H1 and H2.
 88

a3 = 2a1 + a 2 = 2(2) + 4 = 8 = 2 3  

a 4 = 2a 2 + a3 = 2(4) + 8 = 16 = 2 4  
a5 = 2a3 + a 4 = 2(8) + 16 = 32 = 2 5
 
Based on the fact that each term can be written
  as 2 to a power we conjecture that the
general formula of the given sequence is: a n = 2 ∀ n ≥ 0
n

At this point the lecturer stopped writing on the board and pointed out that mathematical
induction can be used to prove that the general formula represents the sequence. After
asking pertinent questions as to how one should go about proving this conjecture it
became clear that the majority of students in the class did not even know how to start
such a proof. Some of the reasons advanced for this are that in their previous experience
with ordinary mathematical induction everything that needed to be proved appeared in
one line and also that they had previously only dealt with one initial value. In this case
however a n = 2 n ∀ n ≥ 0 needs to be proved using a n + 2 = 2a n + a n +1 ∀ n ≥ 0
which appeared on different lines. Most students in the class therefore could not even
start, since they wanted to use only a n = 2 n ∀ n ≥ 0 . Subsequently the lecturer called
Mc Clean (one of the better students) to attempt a proof for the conjecture. The following
is Mc Clean’s proof:

From hypothesis: a k = 2 k where k ≥ 0

Inductive step:
a k +1 = a ( k −1) + 2
= 2a ( k −1) + a ( k −1) + 1

= 2a k −1 + a k
= 2.2 k −1 + 2 k − from hypothesis
= 2 +2 k k

= 2.2 k
= 2 k +1

Mc Clean’s use of (k-1) + 2 shows that he tried to get the subscript in the same form as
the subscript of the a n + 2 in a n + 2 = 2a n + a n +1 (k-1 represents n). This is a clear
indication that he tried to get it to a form that fits the formula, so that he could use it to
expand his hypothesis.

On completion of his proof the lecturer asked Mc Clean to remain at the board. The
lecturer then asked the rest of the class to point out a fundamental flaw in Mc Clean’s
reasoning. No one in the class could point this out. However after a few wild goose
 89

chases by the rest of the students Mc Clean  himself started circling with his finger on
the board around the part he suspected was the problem. The following is that part:
a k +1 = a ( k −1) + 2  

= 2a ( k −1) + a ( k −1) + 1  

= 2a k −1 + a k  
= 2.2 k −1 + 2 k − from hypothesis

This indicated that he started to realize that assuming a k = 2 k which then implies
a k −1 = 2 k −1 is where the problem lies. It also perhaps showed his dawning realization
that he started with what he needed to prove i.e. that if the statement holds for k then it
holds for all other integers. The lecturer confirmed his suspicion by stating that one
cannot start with what you are supposed to prove.

Mc Clean as well as the rest of the students also did not consider the consequences of
letting k start at 0. That is in the inductive step where a k +1 = 2a k −1 + a k if k = 0 then
a −1 would have to be considered. As a result of prompting and incisive questioning
from the lecturer this problem was exposed. The lecturer then suggested to the class to
change this to k ≥ 1 . This was completely new to the class, the fact that this restriction
could be changed, but strangely no student questioned this.
It was by now clear to the students that the ordinary method of proof by induction can
not be used to prove the conjecture. The lecturer then explained that a consequence of
this was that a different and stronger form of mathematical induction was required for
such cases.
The above lesson was deliberately started with the given example and in the indicated
manner so that students could see and identify the limitations of ordinary induction.
Subsequently the students were introduced to the principle of strong mathematical
induction.

5.5.4.5 Principle of Strong Mathematical Induction

Let P (n) be a predicate that is defined for integers n, and let a and b be fixed integers
with a ≤ b . Suppose the following two statements are true:
(i) P(a), P(a+1),…and P(b) are all true (basis step)
(ii) For any integer k > b , if P (i) is true for all integers i with
a ≤ i < k (Inductive hypothesis). Then P (k) is true (inductive step)
Then the statement “for all integers, n ≥ a P (n)” is true.

Example 36 (corrected)
Take P(n) to be “ a n = 2 n ”
Choose a = 0, b = 1
 90

Basis step  
P(0): a 0 = 2 0 = 1
 
P(1): a1 = 21 = 2
∴ P(0) and P(1) are true.  

Inductive hypothesis  
Let k be an integer with k > 1
Suppose that for all integers i with 0 ≤ i < k , ai = 2 i

Inductive step [show that P(k) is true, that is a k = 2 k ]

Now a k = 2a k − 2 + a k −1 , since 0 ≤ k − 2, k − 1 < k
It follows that a k −1 = 2 k −1 and a k − 2 = 2 k − 2
Hence
( )
a k = 2 2 k − 2 + 2 k −1
= 2 k −1 + 2 k −1
= 2.2 k −1
∴ ak = 2 k
This shows that P(k) is true .
So we conclude that a n = 2 n ∀ n ≥ 0

5.5.4.6 Proving a property of a sequence using strong mathematical induction

Example 38
Define a sequence a1 , a 2 , a3 , ... as follows:
a1 = 0 ; a 2 = 2
a k = 3a ⎢ k ⎥ + 2 ∀ int egers k ≥ 3
⎢2⎥
⎣ ⎦

(a) Find the first seven terms of the sequence

(b) Prove that a n is even for each integer n ≥1
 91

Solution  
(a)
a1 = 0  
a2 = 2  
a3 = 3. a ⎢ 3 ⎥ + 2 = 3.a1 + 2 = 3(0) + 2 = 2
⎢2⎥
⎣ ⎦  
a 4 = 3. a ⎢ 4 ⎥ + 2 = 3. a 2 + 2 = 3(2) + 2 = 8
⎢2⎥
⎣ ⎦

a5 = 3. a ⎢ 5 ⎥ + 2 = 3. a 2 + 2 = 3(2 ) + 2 = 8
⎢2⎥
⎣ ⎦

a 6 = 3. a ⎢ 6 ⎥ + 2 = 3. a3 + 2 = 3(2) + 2 = 8
⎢2⎥
⎣ ⎦

a 7 = 3. a ⎢ 7 ⎥ + 2 = 3. a3 + 2 = 3(2) + 2 = 8
⎢2⎥
⎣ ⎦

(b) Let P(n): a n is even

Choose a = 1, b = 2

Basis step
P(a): a1 = 0 which is even
∴P(a) is true.
P(b): a 2 = 2 , which is even
∴ P(b) is true.

Inductive hypothesis
Let k be an integer with k > 2 and suppose that ai is even for all integers i
with 1 ≤ i < k

Inductive step
We know a k = 3. a ⎢ k ⎥ + 2 and k ≥ 3
⎢2⎥
⎣ ⎦

⎧k
if k is even
⎢ k ⎥ ⎪⎪ 2
Remember: =
⎢⎣ 2 ⎥⎦ ⎨ k − 1
⎪
⎪⎩ 2 if k is odd
k k −1 ⎢k ⎥
And 1 ≤ < k and 1 ≤ < k , therefore 1 ≤ ⎢⎣ 2 ⎥⎦ < k
2 2
From the induction hypothesis: a ⎢ k ⎥ is even
⎢2⎥
⎣ ⎦

So 3. a ⎢ k ⎥ is even since [odd × even = even]

⎢2⎥
⎣ ⎦
 92

And so a k = 3. a ⎢ k ⎥ + 2 is also even [since

  it is the sum of two even numbers]
⎢2⎥
⎣ ⎦
 
This shows that P(k) is true
 
We conclude therefore that a n is even for all integers n ≥ 1
 
After the above examples were discussed and analyzed the following similar exercise was
given to students.

Example 39
Suppose b1 , b2 , b3 , ... is a sequence defined as follows:
b1 = 3, b2 = 6, bk = bk − 2 + bk −1 ∀ int egers k ≥ 3
Prove that 3 / bn ∀ int egers n ≥ 1

The following is the solution of Nondumo Masixolo a member of the experimental group
of students:

Solution
Take P (n) to be the statement 3 / bn ∀ int egers n ≥ 1
Choose a = 1 and b = 2

Basis step
P(1): b1 = 3 [3 × 1 = 3]
P(2): b2 = 6 [3 × 2 = 6]
∴P(1) and P(2) are true

Inductive hypothesis
Let k be any integer with k ≥ 2 and suppose ∀ int egers i with 1 ≤ i < k , 3 / bi is true

Inductive step
bk = bk − 2 + bk −1 ∀ int egers k ≥ 3
Now k ≥ 3 = 2 + 1, ∴ k − 2 ≥1 and k − 1 ≥ 2
Since 1 ≤ k − 2, k − 1 < k it follows by inductive hypothesis that 3 / bk −1 and 3 / bk − 2
Then
bk −1 = 3l for some int eger l
bk − 2 = 3v for some int eger v
∴ bk = 3l + 3v
= 3(l + v )
Since l+v is an integer bk = 3 × some int eger and ∴ 3 / bk
 93

Discussion of Masixolo’s solution  

The part of the solution k ≥ 3 = 2 + 1, ∴ k − 2 ≥1 and k − 1 ≥ 2 shows that the student
knew that what is required is to show that the  subscripts k-2 and k-1 is between 1 and k in
order to use his inductive hypothesis. That is he needs to show that the formula holds for
 
k-1 and k-2 in order to prove that it holds for n = k. This in turn is an indication that he
knew that in order to prove it for bk he needed
  to show that 3 / bk −1 and 3 / bk − 2 .
He was also aware of the fact that one can prove 3 divides an integer by showing that it is
3 × some int eger . The conclusion one can draw from the above solution then is that the
student is cognizant of the requirements of such a proof.

Example 40
Prove that any integer greater than 1 is divisible by a prime number.

Solution
Let P (n) be the divisibility property i.e. P (n): n is divisible by a prime number.
Choose a = 2 = b

Basis step
P (a): 2/2 and 2 is prime
Therefore the divisibility property holds for n = 2

Inductive hypothesis
Let k be an integer with k > 2 and suppose that i is divisible by a prime number for all
integers 2 ≤ i < k

Inductive step (proof that k is divisible by a prime)

Now if k is prime then k/k and so k is divisible by a prime number, namely k itself.
If k is composite, then k = r.s where 1 < r < k and 1 < s < k by the composite
number lemma.
By the inductive hypothesis, ∃ a prime p such that p/r and since k = r.s, it follows that
r/k.
Now we have p/r and r/k and so p/k by the transitivity of divisibility.
∴ Regardless of which case holds, there is always a prime number which divides k.
We conclude that every integer >1 is divisible by a prime number.

On completion of this section students were exposed to similar exercises. With some of
these exercises the lecturer asked selected students to explain their solutions to the class.
This was done to determine their level of understanding and also to eliminate reasoning
errors. A pre- and post-test on strong mathematical induction34 was administered. Both of
these tests dealt with recursive sequences. The pre-test was administered immediately
after the lesson on strong mathematical induction that dealt with recursive sequences was
completed, whereas the post-test was done after reasoning errors were addressed.

34
See Appendices I1 and I2.
 94

  the ones stated our study was concluded

Although other topics were dealt with besides
with the above topic.
 

 
 95

 
CHAPTER 6

 
RESULTS OF THE STUDY [PRESENTATION AND DISCUSSION]

6.1 Introduction  

 
In the following section we will present the different assessment instruments and
memoranda and then discuss and analyze our data and findings. We will look at logic
puzzles and proofs separately. For both logic puzzles and proofs we will analyze the
pre-test first and immediately after that the post-test for the same instrument will be
analyzed.

6.2 Logic puzzles

6.2.1 Puzzle I (pre – test)

Compound and conditional statements

In the back of an old cupboard you discover a note signed by a pirate famous for his
bizarre sense of humour and love of logical puzzles. In the note he wrote that he had
hidden treasure somewhere on the property. He listed five true statements (a – e below)
and challenged the reader to use them to figure out the location of the treasure.

a) If this house is next to a lake, then the treasure is not in the kitchen.
b) If the tree in the front yard is an elm, then the treasure is in the kitchen.
c) This house is next to a lake.
d) The tree in the front yard is an elm or the treasure is buried under the flagpole.
e) If the tree in the backyard is an oak, then the treasure is in the garage.

Where is the treasure hidden?

The solution to the puzzle consists of three steps namely:

1. Using statement (a) and (c) we can conclude that the treasure is not in the
kitchen. – modus ponens (this is also known as affirming the consequent)
2. Using statement (b) and the conclusion from 1 we can conclude that the tree in
the front yard is not an elm. – modus tollens ( denying the antecedent)
3. From (d) and the conclusion of 2 we get the final conclusion that the treasure is
buried under the flagpole. – disjunctive syllogism

So in order to solve the puzzle the student had to make three connections between and
from statements.

A memorandum consisting of the above solution was used to assess the students’
answers.
 96

6.2.1.1 Analysis of assessment of puzzle I of  Experimental Group

26 Students completed the puzzle. Thirteen   students (50% of the total number of
students) came to the correct conclusion. However 5 of these students were repeating the
 
course, so effectively only 8 came to the correct conclusion. Six of these students (23%
of the total number of students) gave a complete and correct argument that supported
 
their conclusion. The other 2 gave no supporting arguments for their conclusion. For
example one stated “The oak normally grows next to an ocean, so it can’t be on the
garage”. This statement cannot be deduced from any of the given statements. So it
appears as if the student tried to eliminate some of the possibilities by providing some
motivation for doing so, although in this case the motivation did not make sense.

Thirteen students (50% of the total number of students) came to the wrong conclusion. Of
these students one had no conclusion at all. Four of them made the connection that since
the house is next to a lake it implies that the treasure is not in the kitchen. This shows that
they have made the connection between statements (a) and (c). They however did not use
this conclusion (the treasure is not in the kitchen) to make the other two connections and
therefore came to the wrong conclusion. An example of this can be seen in the following
student’s answer “The treasure is in the garage. Firstly this house is next to the lake so
the treasure is not in the kitchen. The treasure cannot be buried under the flagpole. So
it’s in the garage. Statement (c) says the house is next to a lake and from statement (a)
that means the treasure is not in the kitchen.”

Eight of these students ( representing 31% of the total number of students) gave muddled
reasoning (muddled reasoning - meaning that there was no structure to their argument
i.e. they did not show connections between facts or from facts) and came to conclusions
that cannot be derived from the given statements, for example Andiswa Qosho states “It
is said that if the tree in the backyard is an oak tree then the treasure is in the garage
because a treasure cannot be inside the house if the tree in the back of the yard is an
oak.” Zinnia Williams came to the conclusion that the treasure is not on the property. The
reason advanced for this conclusion is “Neither is the treasure in the front and backyard,
cause there can either be an elm or an oak tree” Siphokazi Ncwaiba stated that since
statements a, b and e use the word if and since if is a keyword one cannot be certain if the
treasure is in the kitchen, garage or under the flagpole. She goes on to state that
statements c and d are more conclusive since it states what is the case. After all this
reasoning the student comes to the conclusion that the treasure is either in the kitchen or
under the flagpole. Thulani Shabangu’s arguments are also very muddled. He states “The
treasure is hidden in a house next to a lake as stated in (c) but it cannot be in the kitchen
according to clue (a). It is buried under a flagpole in a garage”
A summary of the argument analysis of puzzle I for the experimental group is presented
in table 8.
 97

  group (pre-test)
Table 8: Argument analysis of experimental
No. of connections No. of students
0  7
1 10
 
2 2
3  2

6.2.1.2 Analysis of assessment of puzzle I of Control Group

50 Students participated in the study.

24 students (48% of the total number of students) had a correct answer, whilst 26
(representing 52% of students) had incorrect answers.

The 24 correct answers:

Of the 24 correct answers 8 made 0 connections between statements.
8 made one connection between statements.
5 made 2 connections between statements.
3 made 3 connections i.e. made all the required connections i.e.

The 26 Incorrect answers:

Of the 26 incorrect answers 19 made 0 connections.
5 made 1 connection between statements.
2 made 2 connections between statements.
None made 3 connections between statements i.e. there were no students in this group
that made all the connections.
A summary of the argument analysis of puzzle I for the control group is presented in
table 9.

Table 9: Argument analysis of puzzle I of control group (pre-test)

No. of connections No. of students
0 27
1 13
2 7
3 3

6.2.1.3 Conclusions based on puzzle I

Based on the above analysis I think it is safe to assume that at this point in time the
majority of students do not have the skill to make logical connections between
statements. This skill of making connections is also required to make connections
between mathematical statements when proving theorems or statements. In other words
the students need to be able to do deductive reasoning. So in order to teach them to do
deductive reasoning and hence reasoning abstractly we decided to teach them a course in
logic. Epp states that in order for students to think abstractly, it will require them to learn
to use logically valid forms of argument, to avoid common logical errors, to understand
what it means to reason from definitions, and to know how to use both direct and indirect
 98

 
argument to derive new results from those already known to be true. Since proving forms
such a vital part of mathematics it is crucial that students develop this skill. Hence we
used these puzzles to develop in students the  skill of making logical connections between
and from statements.
 
6.2.2 Puzzle III (post test)
 
Your grandfather who is known for his sense of humour and love of logical puzzles left
you a note. In the note he wrote that he had hidden your birthday present somewhere on
one of his properties. He listed five true statements (a – e below) and challenged you to
figure out the location of the present.

a) If this house is next to a main road, then the present is not in the attic.
b) If there is a swing in the yard then the present is in the study.
c) The yard has a lawn or the present is in the cupboard next to the stove.
d) If the yard has a lawn, then the present is in the attic.
e) The house is next to a main road.

Where is the present hidden?

This puzzle is equivalent to puzzle I which was used as a pre – test. This puzzle therefore
was essentially the same as puzzle I only the context was changed i.e.

1. Both puzzles had 5 statements

2. The 1st , 2nd and 3rd statements of puzzle III contained if – then statements
(conditional statements) (compare to puzzle I where statements 1, 2 and 5 is if –
then statements)
3. Statement 3 of puzzle III was a statement containing OR (compare to puzzle I
where statement 4 contained the OR)
4. Puzzle I and III could be proven using:
(i) modus ponens
(ii) modus tollens and then
(iii) disjunctive syllogism
5. Both puzzles contained only compound statements.

The solution to the puzzle is as follows:

1. Using statements (a) and (e) the conclusion is that the present is not in the attic –
modus ponens
2. Using (d) and the conclusion of 1 we get that the yard does not have a lawn –
modus tollens
3. From (c) and the conclusion of 2 we get the final conclusion that the present is in
the cupboard next to the stove – disjunctive syllogism.
So in order to solve the puzzle the student had to make three connections between
and from statements.
 99

  to solve the puzzle:

Alternatively the rules of inference can be used

  the rules of inference:

Using variables with valid argument forms with
Let: p = the house is next to a main road
q = the present is not in the attic  
r = there is a swing in the yard
 
s = the present is in the study
t = the yard has a lawn
u = the present is in the cupboard next to the stove
The statements a to e then become:
(a) p ⇒q
(b) r ⇒s
(c) t ∨ u
(d) t ⇒ ~q
(e) p

Solution:
1. p ⇒ q - from (a)
p - from (e)
∴ q - modus ponens

2. t ⇒ ~ q - from (d)
q - from (1)
∴ ~ t - modus tollens

3. t ∨ u - from (c)
~ t - from (2)
∴ u - disjunctive syllogism
∴ the present is in the cupboard next to the stove

A marking memorandum consisting of the above solutions was used to assess the
students’ answers.
This puzzle was given 6 weeks after the first puzzle.

6.2.2.1 Analysis of assessment of puzzle III (post – test) of experimental group

26 Students attempted the puzzle. Two of the 26 students were repeating the course and
therefore will not be taken into consideration.

Two of the 24 students came to the incorrect conclusion and 22 (92% of the total number
of students) came to the correct conclusion. However one of these students started
midway through the first term and therefore was not present when compound statements
was done in class. The other student, Andiswa Qosho, tried to use variables with the rules
of inference, but only succeeded in rewriting the statements in symbol form. She could
 100

 
not connect the arguments using the rules of inference. So once again she gave a muddled
argument.
 
Seventeen of the students used variables and the rules of inference with valid statement
 
forms to get to the result. One of these students came to the correct conclusion but used
the wrong rule of inference. Another student made a mistake in assigning the variables
 
and therefore had to make use of more than three steps to get to the conclusion.

Four of the students used the statements without variables, but still came to the correct
conclusion. One of these four students came to the correct conclusion although part of his
reasoning was based on an erroneous deduction. He states: “The yard has a lawn but this
house is also next to a main road, therefore the present can’t be in the attic. ∴ The
present is in the cupboard next to the stove.”

Of the four students that initially could not progress beyond the first connection (with
puzzle I) three now got it completely correct whilst the fourth one did not do puzzle III.
Three of the eight that gave muddled arguments with puzzle I, now gave a complete
argument and came to the correct conclusion. Of this eight four did not do puzzle III and
one ceased her studies.
A summary of the argument analysis of puzzle III for the experimental group is presented
in table 10.

Table 10: Argument analysis of puzzle III of experimental group (post-test)

No. of connections No. of students
0 1
1 1
2 5
3 17

6.2.2.2 Analysis of assessment of puzzle III of control group

49 Students participated in the study.

31 of the 49 had correct answers (this represents 63% of the total number of students)
18 of the 49 had incorrect answers (37% of the total number of students)

The 31 correct answers:

3 made 0 connections between statements
14 made 1 connection between statements
5 made 2 connections between statements
9 made 3 connections between statements

The 18 incorrect answers:

10 made 0 connections between statements
4 made 1 connection between statements
4 made 2 connections between statements
0 made 3 connections between statements.
 101

 
A summary of the argument analysis of puzzle I for the control group is presented in
table 11.
 
Table 11: Argument analysis of puzzle III of control group (post-test)
 
No. of connections No. of students
0  
13
1 18
2 9
3 9

6.2.3 Comparison of answers of puzzle I and puzzle III

Table 12 gives a comparison of the attempts of some of the students of the experimental
group. This was done in order to show their increasing ability to argue deductively.

Table 12: Comparison of answers of puzzle I and puzzle III

Puzzle I Puzzle III
Marsha MacMahon
- A statement was made that the house is From statement a and e, I gathered that the
next to a lake present is not in the attic. With the present
- Another statement was made thereafter not being in the attic the yard does not have
that the tree in the front yard is an elm a lawn;
∴ The treasure is in the kitchen. ∴ statement d is wiped out;
∴ I can say that the present is in the
cupboard next to the stove, because the first
half of statement c is not true. There is no
facts to make statement b correct
∴ its redundant
∴ In my opinion the present is in the
cupboard next to the stove.

Bernarain Mvondo
The treasure is hidden in his boat because (1) p ⇒q
he says this house is next to a lake which (2) r ⇒s
means that the boat is ashore;
(3) t ∨ u
The tree in the front yard represents the t ⇒ ~q
symbol of the boat who carries the (4)
flagpole. (5) p
And kitchen, garage, flagpole are the words (1) & (5) q (6)
used by pirates. (6) & (4) ~t (7)
(7) & (3) u
∴ the present is in the cupboard next to the
stove
Zukile Roro
The treasure is in the garage. Statement (e) says the house is next to a
Firstly this house next to the lake so the main road. Using this fact we can conclude
 102

treasure is not in the kitchen. The treasure   that the present is not in the attic as it is
can not be buried under the flagpole. stated in statement (a)
So it’s in the garage.  
Statement (c) says the house is next to a Using statement (d) I can say the yard has
lake and from statement (a) that means the   no lawn because from the above paragraph
treasure is not in the kitchen. I know that the present is not in the attic.
 
So this yard has no lawn.

Statement (c) states that the yard has a lawn

or the present is in the cupboard next to the
stove.

I have already proved that the yard has no

lawn and now I can conclude that the
present is in the cupboard.

The present is hidden in the cupboard next

the stove.

Phumla Thafeni
The treasure is in the kitchen because this The present is in the cupboard next the
house is next to a lake and the tree in the stove.because statement 5 says that the
front yard is an elm house is next to the main road and that
The fact that the house is next to the lake means the present cannot be found in the
no one can guess that the treasure might be attic according to statement 1.
buried under the flagpole. And it is Therefore the yard does not have a lawn.
impossible that the treasure can be buried 1. p ⇒ ~ q
because of the lake that is next to the 2. r ⇒ s
3. w ∨ t
house. Probably inside the kitchen.

4. w ⇒ q
5. p
6. ~ q 1,5 modus ponens
7. ~ w 4,6 modus tollens
8. t 3,7 disjunctive syllogism
9. p ∧ t 5,8 conjunctive addition

Siphokazi Ncwaiba
The elm tree is in the front yard which I found out that the present is in the
means the kitchen is in front too. The lake cupboard next to the stove.
is behind the house so the treasure is not at It is stated that the house is next to a main
the lake. So this house next to the lake road so it’s not in the attic. Then again it
which means the kitchen is out of the says if the yard has a lawn, it is in the attic.
picture. The yard doesn’t have a lawn so its not in
The tree is an elm right. the attic.
In my own thinking I found out that This is a contradiction.
 103

statements a; b; e only state if the items are   But I see that it’s not in the study because
this then its there, if being a keyword so we don’t know if the house has a swing.
you not sure if its in the kitchen, garage or   So we left with the cupboard next to the
under the flagpole. stove.
While c; d state what it is.   1. p ⇒ ~ q
1. kitchen – was in the lead, we know
  2. r ⇒ s
dat the tree is an elm and that its
next to the lake. 3. t ∨ u
2. flagpole – 4. t ⇒ q
my conclusion comes to the kitchen or 5. p
flagpole.
6. ~ q 1 and 5 mod us ponens
7. ~ t 4 and 6 mod us tollens
8. u 3 and 8 disjunctive syl log ism

Andiswa Hazel Qosho

The treasure is in the garage because it is If there is a swing in the yard then the
said that if the tree in the backyard is an present is in the study. I think this is where
oak then the treasure is in the garage the present is because if there is a swing in
because a treasure can not be inside the the yard then the present cannot be outside
house if the tree in the backyard is an oak the house cause the swing can be the
enemy and modus pones it is said that
p⇒q
∴q
Which implies that if the swing is in the
yard then the present is in the study.
1. p ⇒ q
2r ⇒s
3. t ∨ w
4. t ⇒ ~ q
5. ∴ p
1 and 5

6.2.4 Conclusions based on the analysis of puzzle I and puzzle III

The puzzles show how students learn to make connections between sentences and then
deduce conclusions, which are vital tools in proving mathematical statements. 92% of
students of the experimental group came to the correct conclusion with the post-test,
whereas only 50% of these students came to the correct conclusion with the pre-test.
Therefore there was a dramatic improvement in the number of students that could solve
the puzzle. This improvement occurred after the students had received teaching on the
logic of compound statements and the logic of quantified statements. Furthermore prior
to receiving teaching the majority of these students could only make one or zero
 104

connections between statements, whereas   after receiving teaching the majority of

students made all three of the necessary deductions. Compare this to the control-group
  and post-test could only manage to make
where the majority of students in both the pre-
one or zero connection between and from statements. The control-group therefore
showed very little improvement between  pre- and post-test although they received
extensive teaching on Differential Calculus in the interval between the pre- and post-test.
 
This surely then must be an indication that knowledge of the logic of compound and
quantified statements have helped to improve the ability of the experimental group of
students to make logical connections between these kinds of statements.

It is also imperative that students really understand the implication of every word in a
statement. Regarding this Epp [33] states the following: “In order to evaluate the truth or
falsity of a statement, you must understand what the statement is about. In other words,
you must know the meanings of all terms that occur in the statement.” One of the goals of
the puzzles therefore was to focus the attention of students on the implication of certain
words in a statement. For example the results of the puzzles shows the growing
awareness of the experimental-group of students that words like OR means that at least
one of the two statements is true. The comparison of the two puzzles therefore show that
these students are starting to read with understanding and are starting to realize that in
Mathematics a word has a specific meaning. This reading with understanding is
absolutely crucial in unraveling mathematical statements. Hence our teaching was geared
towards enhancing, in students, this ability of reading with understanding.

6.3 Puzzles on knights and knaves

The solution to these puzzles is based on the use of proof by contradiction. In other words
the student has to make an assumption and then show that the given facts contradict the
assumption. This has to be done by making logical connections between facts or making
deductions from facts.

This pre-test of the puzzle was given to the students without giving them any teaching on
proof by contradiction. The puzzle is presented below.

6.3.1 Knights and knaves (pre-test)

The logician Raymond Smullyan describes an island containing two types of people:
knights who always tell the truth and knaves who always lie. You visit the island and are
approached by two natives who speak to you as follows:

A says: B is a knight
B says: A and I are of opposite type.

What are A and B?

 105

Solution  
Suppose A is a knight.
 
Therefore what A says is true - (by defn. of knight)
Therefore B is a knight also – (this is what A said)
 
Therefore what B says is true – (by defn. of knight)
Therefore A and B are of opposite type – (this is what B said)
 
We have the following contradiction: A and B are both knights and A and B are of
opposite type.
Therefore A is not a knight
Therefore A is a knave
Therefore what A says is false (by defn. of knave)
Therefore B is not a knight.
Therefore B is a knave also.
A and B are both knaves.

A marking memorandum consisting of the above solution was used to assess student
answers.

6.3.1.1 Analysis of arguments of students

22 students attempted the puzzle.

15 students could not solve it.
7 students gave valid arguments.

Analysis of the 7 correct arguments:

5 students did the puzzle by contradiction:( Thoriso Seepi; Bernarain Mvondo; Garren
Davidse; Markan Mclean; Thulani Shabangu). These students used the correct method of
argument to get a contradiction. For example the following is Thulani Shabangu’s
argument: “If B is indeed a knight then A is telling the truth, but then A must be a knave.
But knaves always lie, therefore B cannot be a knight and A can never be a knight, since
he said B is the one and B says A and I are of opposite type. Which are two statements
that contradicts and knights never lie therefore they are both knaves.”

Brian Masona used cases to solve the puzzle, but the cases are also based on
contradicting an original assumption. His argument is as follows: “ 1. If A is telling the
truth, that also makes him a knight. But then there would be no reason to lie on B’s part
if he is a knight as well. (wrong)
2. If B is telling the truth, it would mean he is the knight and A is the knave. However this
also means A is telling truth. Therefore this is wrong.
3. If both are lying then A is not opposite to B and B is not a knight, so that makes them
both knaves. Therefore this is true.”
Ashwin Patience did his by process of elimination, but it is still based on contradicting
an original assumption.
 106

Analysis of the 15 incorrect arguments:  

Four students (Geraldo Maasdorp,Busisiwe Qavane, Segodi Evans, Phumla Thafeni) had
some argument but their argument did not lead   to the correct conclusion. For example the
following is Phumla Thafeni’s argument: “ B belongs to the knight people because A has
 
stated that B is a knight. A belongs to the knaves people who always tell lies because the
lie is the opposite of the truth as B stated.” Although Pumla has assumed that the first
 
statement is correct i.e. that A is telling the truth and that therefore A is a knight, she
contradicts herself by stating in the next sentence that A is a knave. This means then that
she did not make the connection between her two statements which could have led to the
contradiction that she needed to prove that A is a knave. The skill to induce a
contradiction and then using this to draw a conclusion is therefore obviously lacking in
Pumla’s reasoning armour at this point in time.

Four students (Yasser Buchana, Marsha Mac Mahon, Masixole Nondumo, Siphokazi
Ncwaiba) came to the correct conclusion but their supporting arguments did not support
their conclusion. For example the following is Marsha Mac Mahon’s argument: “ B is a
knave because A always lies, therefore B is not a knight but a knave. Because B is a
knave he always lies, therefore A and B are of the same type. Therefore A and B are both
knaves.” Marsha does not show how she came to the first assumption “B is a knave
because A is lying”. The rest of her argument however is logical. This shows that
intuitively she knew the answer, but lacks the skills to show how she got to the answer
i.e. the step-by-step reasoning that allows you to connect one fact to another or to deduce
a conclusion from a fact. This step-by-step reasoning is an important skill in proving
mathematical statements.
Some of the student arguments had elements of contradiction in it, but since the students
did not know how to use the contradiction, they did not give a clear and concise argument
to show how they reached their conclusion. For example Siphokazi Ncwaiba’s argument
is as follows: “A states that B is a knight, ok if that’s true it means A is a knave but the
statements says knaves always lie so how do we know that A is telling the truth.
B states that A and B are opposite types which means one is lying and the other one is
telling the truth. So now who is telling the truth and who’s lying. This thing revolves
around the question that says who’s lying and who’s telling the truth.
Conclusion
A is a knave
B is a knave
It comes to this A is stating that B is a knight, if its true it automatically makes A a knave
therefore a knave is a liar how sure are we that he is telling the truth.”
In the very first two lines of Siphokazi’s argument a contradiction arises, but since she
does not recognize this, she does not use it. Since the rest of her argument depends on her
using this contradiction she ends up giving a muddled argument.
A summary of the argument analysis of the pre-test of the knights and knaves puzzle is
presented in table 13.
 107

  knaves pre-test
Table 13: Argument analysis of knights and
No. of connections No. of students
0   11
1 1
 
2 0
3  2
4 0
5 0
6 1
7 1
8 3
9 0
10 0
11 0
12 0

The post-test was presented to students approximately 6 weeks after the pre-test.

6.3.2 Knights and knaves (post - test)

The logician Raymond Smullyan describes an island containing two types of people:
knights who always tell the truth and knaves who always lie. You visit the island and are
approached by two natives C and D but only C speaks.

C says: Both of us are knaves.

What are C and D?

Solution
Suppose C is a knight. Then what C says is true. - (by defn. of knight)
Therefore C and D are both knaves. – (this is what C said)
Therefore C is a knave.
Thus we have a contradiction: C is both a knight and a knave.
Hence our supposition that C is a knight is false.
Hence C is a knave.
Therefore what C says is false. – (by defn of knave)
Hence either C or D is not a knave. (De Morgan’s law)
But C is a knave, hence D is not a knave.
Therefore D is a knight.

6.3.2.1 Analysis of arguments of students

29 Students attempted the puzzle.

7 gave incorrect answers.
10 gave correct answers, but had gaps in their arguments.
12 gave complete and valid arguments.
 108

Analysis of the 7 incorrect answers  

Although the solution of these 7 students were all incorrect, they all attempted to solve
the puzzle by using proof by contradiction,  some with more success than others. For
example Luthando Myeki argues as follows: “The supposition is that both of them are
 
knaves, then both C and D are lying. But I can’t say what C say is a lie, because D does
not say anything. Therefore there is a contradiction because C spoke and D did not. Then
 
the supposition is false, because at least one of them must be a knight. Therefore D is a
knight.
Luthando argues that a contradiction is induced because C spoke and D did not. It is
obvious therefore that he does not fully understand when and how a contradiction arises.
The rest of his argument however is logically correct since he correctly negates the
inferred AND with an OR. In other words he has correctly applied De Morgan’s law i.e.
since he initially assumed that both C and D are knaves, he correctly assumes the
opposite that either C or D must be a knight after the “contradiction”.
[Andiswa Qosho; Brian Masona; Luthando Myeki; Carmen Williams; Phumla Thafeni;
Masixolo Nondumo; Ashwin Paulse]

Analysis of the 10 arguments with gaps

Most of these students correctly used contradiction to conclude that C is a knave, they
however did not make use of this fact to advance their argument and hence to conclude
that D is a knight. As a result there is a gap in their argument, since they do not show how
they have reached their conclusion that D is a knight. For example Siyabonga Maki
argues as follows: “Suppose that C is a knight. Therefore C will always tell the truth.
Therefore both C and D are knaves and we have a contradiction because C can not be a
knight and a knave. Hence our supposition is false. Therefore C is a knave and D is a
knight.”

Analysis of the 12 correct arguments

These students presented arguments that contained all the essential steps of an acceptable
solution. For example Dionisio Nunes argues as follows: “Suppose C is a knight. Then
what C says is true. Therefore C in particular is a knave. But this is a contradiction.
Therefore our supposition that C is a knight is false. Therefore C is a knave. Then what C
says is false. Therefore C or D is not a knave. We have already concluded that C is a
Knave. Therefore D is not a knave. Hence D is a knight. Therefore C is a knave and D is
a knight.”
A summary of the argument analysis of the post-test of the knights and knaves puzzle is
presented in table 14.
 109

  knights and knaves puzzle

Table 14: Argument analysis of post-test of
No. of connections No. of students
0  4
1 0
 
2 1
3  0
4 1
5 3
6 4
7 3
8 5
9 3
10 3
11 2

6.3.3 Conclusions based on the analysis of knights and knaves puzzles

In both the pre and post –test the percentage of students that gave completely valid
arguments are less than 50% of the total number of students that attempted the puzzles.
Add to this the fact that almost 70% of students could not use contradiction to prove their
argument in the pre-test then this surely is an indication that students do not find proof by
contradiction an easy exercise. The fact that the majority (more than 70%) of students
could induce a contradiction in the post-test however is testimony to a major
improvement in proving skills as far as proof by contradiction is concerned. A
comparison of some student answers to the pre- and post-test of the knights and knaves
puzzles are presented in table 15.

Table 15: Comparison of answers to pre- and post-test of knights and knaves
puzzles
Pre-test Post-test
Marsha MacMahon
B is a knave because A always lies. Suppose C is a knight. Then what he is
Therefore B is not a knight but a knave. saying is the truth. Therefore he is a knave.
Because B is a knave he always lies. Which is a contradiction because C cannot
Therefore A and B are of the same type. be both a knight and a knave. Therefore my
Therefore A and B are both knaves. supposition is false. So therefore at least
one of them is a knight. So suppose C is a
knave. Then what he is saying is a lie.
Therefore they are not both knaves. But
one of them are a knave. But because C is
lying, he is a knave. Therefore D is a
knight. C - knave; D – knight.
Phuti Senyatsi
A and B are both knights. We could say Suppose C is a knight. Therefore what C
that A’s statement is untrue, but we cannot says is true. However, C says both him and
 110

in anyway say that B’s statement is untrue   D are knaves. There is therefore a
because it is too ambiguous, therefore it contradiction because C cannot be a knight
might have some truth to it. Therefore,   and a knave at the same time. Therefore
because there is a possibility that B’s our supposition is wrong. Suppose C is a
statement might have some truth, that   knave. Therefore what C says is false. This
makes him a knight. This then makes A’s means that there is at least one of them who
 
statement also true. A is therefore a knight are a knight. Therefore D is a knight and C
as well. is a knave.
Segodi Evans
B said that A and him are of opposite type Suppose C is a knight. So according to our
which means that B is right because statement what C says is true. Therefore C
according to the statement the two natives is a knave. But C cannot be both knight and
are completely different. A said that B is a knave therefore we have a contradiction.
knight. A could be wrong because on the So our supposition is false. Therefore C is a
statement Raymond said that between the knave. So what C says is not true. Both of
two natives one is a liar. A is the one them are not knaves. Therefore D is a
whose statement is true in either way. So it knight.
means A is the one who always tells the
truth. While B’s statement is not clear, he
could be the one who always lies.
Zukile Roro
B is a knight and A is a knave. A is the one Suppose C is a knight. Then what C says is
who always lie. He introduces B first true. Therefore both of them are knaves.
instead of himself. B says they are of But that is impossible, C cannot be a knight
opposite type meaning that if B is the and a knave. So that is a contradiction.
knight then A is the knave. That means my supposition that C is a
knight is false. Therefore C is a knave. If
C is a knave then one of them has to be a
knight. Therefore D is a knight.
Sigqibo Lande
B is a knight given that A tell us that B is a Suppose C is a knight. Therefore C tells the
knight which means both A and B are truth. Therefore C is telling a truth when C
knights, but they differ in gender. says that both of them are knaves. Thus
there is a contradiction that you can’t tell a
truth and be a knave.(taking what C says).
Therefore my supposition is false.
Therefore C is a knave and D is a knight.
(from the opposite of both). There is at
least one knave and knight. I’ve already
proven that C is a knave. That’s why I say
that D is a knight.
 111

6.4  
Arguments with Quantified statements

6.4.1 Pre-test  

 
Students had to rewrite the following statements in formal Mathematical language
 
Question A
1. No birds, except ostriches, are nine feet high.
2. There are no birds in this aviary that belong to anyone but me
3. No ostrich lives on mince pies.
4. I have no birds less than nine feet high.
Question B
1. All writers who understand human nature are clever.
2. No one is a true poet unless he can stir the hearts of men.
3. Shakespeare wrote Hamlet.
4. No writer who does not understand human nature can stir the hearts of men.
5. None but a true poet could have written Hamlet.

Solution
Question A
1. ∀ x, if x is not an ostrich, then x is less than nine feet tall.
2. ∀ x, if x is a bird in this aviary, then x is a bird which belongs to me.
3. ∀ x, if x is an ostrich, then x does not live on mince pies.
4. ∀ x, if x is a bird that is less than nine feet high, then x does not belong to me.

Question B
1. ∀ x, if x is a writer who understands human nature, then x is clever.
2. ∀ x, if x cannot stir the hearts of men then x is not a true poet.
3. Shakespeare wrote Hamlet.
4. ∀ x, if x is a writer who does not understand human nature, then x cannot stir the
hearts of men.
5. ∀ x, if x is not a true poet, then x could not have written Hamlet.

6.4.1.1 Analysis of student answers of Quantified Statements

Question A
The majority of students did not rewrite number 1 correctly because they did not interpret
the word “except” correctly and hence did not use the negation “not an ostrich” in their
answer. For example Phuti Senyatsi writes: “ ∀ birds x, if x is nine feet high, then x is an
ostrich”
Numbers 2 and 3 was answered correctly by most students.
Number 4 was answered incorrectly by most students, since again they failed to use the
universal quantifier (negation) for “no birds”. For example Siphokazi Ncwaiba writes:
“ ∀ birds x, if x belongs to me, then x is nine feet high.”
 112

 
Question B
 
The majority of students rewrote number 1 correctly.
The words “no one” and “unless” caused most students to write number 2 incorrectly as
they failed to use negation. For example Yasser   Buchana writes: “ ∀ poets x, if x is a true
poet, then x can stir the hearts of men.”
 
Most students did not realize that number 3 could not be rewritten in formal language.
The majority of students could not interpret “no writer” and “does not understand”
correctly and hence rewrote it incorrectly. An example of this can be seen in Carmen
Williams’ answer: “ ∀ writers x, if x understands human nature, then x can stir the hearts
of men”
Very few students could rewrite number 5 in formal language as again they had problems
with “none” as they did not interpret this to mean negation. Phuti Senyatsi’s answer
illustrates this: “ ∀ poets x, if x wrote Hamlet, then x is true.”

The above analysis shows that students had problems with sentences that have the
negation at the beginning. Students also had problems with words like none, unless,
except, etc. and could not rewrite it properly in formal mathematical language as they did
not interpret the meaning of these words correctly. We therefore decided on an
intervention strategy that would allow the students to correctly interpret and hence
rewrite such statements correctly. The strategy included discussion and more practice
exercises on the above type of exercises.
A summary of the argument analysis of the pre-test of the quantified statements are
presented in table 16 and 17.

Table 16: Argument analysis of quantified statements (question 1) pre-test

Score No. of students
0 2
1 5
2 19
3 5
4 0

Table 17: Argument analysis of quantified statements (question 2) pre-test

score No. of students
0 1
1 12
2 15
3 3
4 0
5 0
 113

6.4.2 Post-test quantified statements  

The following were used as a post-test:  

Rewrite the following statements in formal language
 
1. No bank closes before 3:30 unless it is a small bank.
 
2. No shark eats plankton unless it is a whale shark.
3. Students never study unless they have to prepare for a test.
4. None but a true gentleman will offer his seat to a lady on a bus.
5. None but a brave soldier will fight in a war

Solution
1. ∀ x, if x is not a small bank then x is a bank that does not close before 3:30.
2. ∀ x, if x is not a whale shark then x is a shark that does not eat plankton.
3. ∀ x, if x is a student that does not have to prepare for a test then x is a student
who never studies.
4. ∀ x, if x is not a true gentleman, then x is a gentleman who will not offer his seat
to a lady on a bus. OR ∀ x, if x offers his seat to a lady on a bus, then x is a true
gentleman.
5. ∀ x, if x is not a brave soldier then x is soldier who will not fight in a war. OR
∀ x, if x fights in war, then x is a brave soldier.

6.4.2.1 Analysis of student answers of Quantified Statements

The majority of students performed quite well in the post-test. The ability to translate
statements from informal to formal language and vice versa is a requisite skill necessary
for forming conclusions from arguments. Furthermore some regard the universal
conditional statement as the most important form of statement in mathematics. They are
of the opinion that familiarity with statements of this form is essential if one is to learn to
speak mathematics. It was therefore imperative for students to be well versed in this skill.
A summary of the argument analysis of the post-test of the quantified statements is
presented in table 18.

Table 18: Argument analysis of quantified statements (post-test)

score No. of students
0 1
1 0
2 4
3 11
4 8
5 5
 114

6.4.3  
Arguments with Quantified statements (forming conclusions)

The reason why these quantified statements  with conclusions were done was to help
students to learn how to connect statements in the correct order to form a conclusion.
This is a very necessary and important skill  in the proving of mathematical statements,
since in proving mathematical statements one needs first of all to be able to understand
 
where and how to start the proof. Consequently one needs to connect each statement
logically with the previous one until a logical conclusion is reached.

6.4.3.1 Puzzle I (pre – test)

Reorder the premises in the following argument to make it clear that the conclusion
follows logically. It may be helpful to rewrite some of the statements in if – then form
and to replace some statements by their contrapositives.

1. When I work a logic example without grumbling, you may be sure it is one I
understand.
2. The arguments in these examples are not arranged in regular order like the ones I
am used to.
3. No easy examples make my head ache.
4. I can’t understand examples if the arguments are not arranged in regular order like
the ones I am used to.
5. I never grumble at an example unless it gives me a headache.

∴ These examples are not easy.

Solution
2. The arguments in these examples are not arranged in regular order like the ones I
am used to.
4. If the arguments are not arranged in regular order like the ones I am used to, then
I can’t understand the examples.
1. If I do not understand a logic example, then I grumble at it.
5. If I grumble at an example then it gives me a headache.
3. If an example gives me a headache, then it is not an easy example.
∴ These examples are not easy.

Alternatively formal language and the rules of inference can be used:

Let:
Q = The arguments in these examples are not arranged in regular order like the
ones I am used to.
P = Do not understand these examples
R = grumble at an example
S = Example give me a headache
T = It is not an easy example
 115

The statements in the correct order in formal  language become:

2. Q
4. Q ⇒ P  
1. ~R ⇒ ~P
 
5. R ⇒ S
3. S ⇒ T  
6. P (from 2 & 4 – modus ponens)
7. R (from 1 & 6 – modus ponens)
8. S (from 5 & 7 – modus ponens)
9. T ((from 3 & 8 – modus ponens)
∴ These examples are not easy

Note: Marks was allocated whenever two statements were given in the correct
order, even if the other statements were not in the correct order.
Marks was subtracted if a statement was not rewritten properly in if-then
form.

6.4.3.2 Analysis of student answers

A total of 20 students attempted the puzzle. Nine students (45% of the total number of
students) gave a complete and valid argument. Eleven students (55% of the total number
of students) presented incorrect arguments.

Fourteen students attempted to rewrite the statements in formal language and thereafter
arrange them in the correct order. Seven of these students had correct answers and seven
gave incorrect answers. The remaining six used the statements as they are. Two of these
supplied a correct answer and four and incorrect answer.
A summary of the argument analysis of the pre-test of the quantified statements (forming
conclusions) is presented in table 19.

Table 19: Argument analysis of quantified statements (forming conclusions) pre-test

No. of connections No. of students
0 6
1 3
2 1
3 1
4 1
5 8
 116

6.4.3.3 Puzzle II (post – test)  

 
Reorder the premises in the following argument to make it clear that the conclusion
follows logically. It may be helpful to rewrite some of the statements in if – then form
 
and to replace some statements by their contrapositives.
 
6. There is no box of mine here that I dare open.
7. My writing-desk is made of rose-wood.
8. All my boxes are painted, except what are here.
9. There is no box of mine that I dare not open, unless it is full of live scorpions.
10. All my rose-wood boxes are unpainted.

∴ My writing-desk is full of live scorpions.

Solution
2. My writing desk is made of rose-wood
5. If my writing-desk is made of rose-wood, then it is unpainted
3. If my boxes are unpainted, then it is here
1. If a box is here, then I dare not open it
4. If I dare not open a box, then it is full of live scorpions
∴ My writing-desk is full of live scorpions

Alternatively formal language and the rules of inference could be used:

Let: P = boxes that I dare open
Q = boxes full of live scorpions
R = boxes that are here
S = boxes made of rose-wood
T = boxes that are painted
U = my writing-desk

The statements in the correct order in formal language become:

2. U ⇒ S
5. S ⇒ ~T
3. ~T ⇒ R
1. R ⇒ ~P
4. ~P ⇒ Q
6. U ⇒ ~T (from 2 & 5 – hypothetical syllogism)
7. U ⇒ R (from 6 & 3 – hypothetical syllogism)
8. U ⇒ ~P (from 7 & 1 – hypothetical syllogism)
9. U ⇒ Q (from 8 & 4 – hypothetical syllogism)
∴ My writing-desk is full of live scorpions.
 117

6.4.3.4 Analysis of student answers  

Twenty five students attempted the puzzle.   No students used formal language and the
rules of inference .Thirteen students (52% of the total number of students) gave a
complete and valid argument. Five students  supplied the correct order for the statements
but rewrote one statement erroneously in if-then form. Three students had the correct
 
order, but made mistakes in the rewriting of two statements. One student had the correct
sequence, but made three errors in rewriting in if-then form. Three students had a
completely incorrect sequence, two of which did not rewrite in if-then form and hence
had a zero score.
A summary of the argument analysis of the post-test of the quantified statements
(forming conclusions) is presented in table 20.

Table 20: Argument analysis of quantified statements (forming conclusions) post-

test
No. of connections No. of students
0 2
1 0
2 2
3 3
4 5
5 13

6.4.4 Conclusions based on the results of Quantified Statements:

Although the improvement from pre- to post-test was not dramatic as far as complete and
valid answers are concerned, the number of students that had the correct sequence with
the post-test did increase significantly ( 88% of the total number of students). However
the results of the post-test seem to suggest that students still have some problems in
rewriting statements in the correct if-then form.

6.5 Proofs

6.5.1 Set Theory

The following pre-test was administered to students:

6.5.1.1 Pre-test
For all sets A, B and C prove the following:

A ∩ ( B ∪ C ) = ( A ∩ B) ∪ ( A ∩ C )
 118

Solution  
Suppose x is a particular, but arbitrarily chosen element of A ∩ ( B ∪ C ) ,
 
then x ∈ A and x ∈ B ∪ C , so x ∈ A and x ∈ B or x ∈ C
hence x ∈ A and x ∈ B or x ∈ A and  x ∈ C
Case 1: x ∈ A and x ∈ B
 
then x ∈ A ∩ B , and so x ∈ ( A ∩ B ) ∪ ( A ∩ C )
Case 2: x ∈ A and x ∈ C
then x ∈ A ∩ C and so x ∈ ( A ∩ B ) ∪ ( A ∩ C )
hence in either case x ∈ ( A ∩ B ) ∪ ( A ∩ C )
so A ∩ ( B ∪ C ) ⊆ ( A ∩ B ) ∪ ( A ∩ C )

Conversely suppose x ∈ ( A ∩ B ) ∪ ( A ∩ C ) , then

x ∈ A ∩ B or x ∈ A ∩ C
Case 1 :( x ∈ A ∩ B )
then x ∈ A and x ∈ B
since x ∈ B , x ∈ B ∪ C , hence x ∈ A and x ∈ B ∪ C and so x ∈ A ∩ ( B ∪ C )
case 2: ( x ∈ A ∩ C )
then x ∈ A and x ∈ C
since x ∈ C , x ∈ B ∪ C , and so x ∈ A and x ∈ B ∪ C , hence x ∈ A ∩ ( B ∪ C )

so in either case x ∈ A ∩ ( B ∪ C )
hence ( A ∩ B ) ∪ ( A ∩ C ) ⊆ A ∩ ( B ∪ C )
hence A ∩ ( B ∪ C ) = ( A ∩ B ) ∪ ( A ∩ C )

NOTE: If students employed a shorter method and it was obvious that the other steps was
implicit in the reasoning of the student, the student was accredited with
marks for these implicit steps.

6.5.1.2 Analysis of student answers

23 students completed the pre-test. Thirteen students ( 57% of the total number of
students) achieved a score of 30% or less and two scored 45% or less. Three of these
students had a zero score, two of which supplied no answer and one (Marsha Mac
Mahon) supplied a muddled answer. Marsha’s attempted solution is as follows:
x ∈ A ∩ B or x ∈ A ∩ C
Case 1: x ∈ C
If x ∈ C then x ∈ A ∩ C but then x ∉ A ∩ B ∴ x ∉ C

Case 2: x ∈ B
 119

If x ∈ C then x ∈ A ∩ B but then x ∉ A  ∩ C ∴ x ∉ B

Case 3: x ∈ A
 
If x ∈ A then x ∈ ( A ∩ B ) ∪ ( A ∩ C )
∴x∈ A  
∴ A ∩ ( B ∪ C ) = ( A ∩ B) ∪ ( A ∩ C )  

It is clear that the student does not know how and when to use cases. In her use of cases
she attempted to show that x is not an element of sets A, B and C. This is an indication
that she is not aware of what is required i.e. that she must prove
A ∩ ( B ∪ C ) ⊆ ( A ∩ B ) ∪ ( A ∩ C ) and
conversely ( A ∩ B ) ∪ ( A ∩ C ) ⊆ A ∩ ( B ∪ C ) .
She also misinterprets intersection since she states “ If x ∈ C then x ∈ A ∩ C ”. She
is therefore definitely not cognizant of the basic definitions of set theory.
The majority of the other students in this group were aware that they were supposed to
show A ∩ ( B ∪ C ) ⊆ ( A ∩ B ) ∪ ( A ∩ C ) and conversely
( A ∩ B ) ∪ ( A ∩ C ) ⊆ A ∩ ( B ∪ C ) . They however had gaps in their arguments or
they used inappropriate cases leading to erroneous arguments.

Six students (26% of the total number of students) attained a score between 55% and
80%. Siyabonga’s proof is as follows: Case 1: x ∈ A ∩ ( B ∪ C )
Then x ∈ A and x ∈ B or x ∈ C therefore x ∈ ( A ∩ B ) or ( A ∩ C )
and hence x ∈ ( A ∩ B ) ∪ ( A ∩ C )
∴ A ∩ ( B ∪ C ) ⊆ ( A ∩ B) ∪ ( A ∩ C )
Case 2: x ∈ ( A ∩ B ) ∪ ( A ∩ C )
then x ∈ A ∩ B or x ∈ A ∩ C
then ( x ∈ A and x ∈ B ) or ( x ∈ A and x ∈ C )
therefore x ∈ A ∩ B or x ∈ A ∩ C
therefore x ∈ A ∩ ( B ∪ C ) and hence ( A ∩ B ) ∪ ( A ∩ C ) ⊆ A ∩ ( B ∪ C )
∴ A ∩ ( B ∪ C ) = ( A ∩ B) ∪ ( A ∩ C )
Siyabonga made inappropriate use of cases and has a circular argument in his second
case. He does however use the appropriate definitions correctly and he seems to
understand that he needs to show A ∩ ( B ∪ C ) ⊆ ( A ∩ B ) ∪ ( A ∩ C ) and conversely
( A ∩ B) ∪ ( A ∩ C ) ⊆ A ∩ ( B ∪ C ) .
Three of the students of this group of six applied the rules of inference in their reasoning.
They showed in their margins the rule of inference that they have applied to make a
deduction. (Nyirenda Pereka; Evans Segodi; Markan McLean). It is notable that these
students were among the highest scorers. The following is the solution of such a student
namely Evans Segodi:
 120

Suppose x ∈ A ∩ ( B ∪ C ) , therefore x ∈  A and x ∈ B ∪ C

x ∈ A and ( x ∈ B or x ∈ C ) but ( x ∈ A and
 
x ∈ B ) or ( x ∈ A and x ∈ C )
Therefore x ∈ ( A ∩ B ) ∪ ( A ∩ C )
∴ A ∩ ( B ∪ C ) ⊆ ( A ∩ B) ∪ ( A ∩ C )  

 
Conversely
Suppose x ∈ ( A ∩ B ) ∪ ( A ∩ C ) then ( x ∈ A ∩ B ) or ( x ∈ A ∩ C )
Case 1: ( x ∈ A ∩ B )
Then x ∈ A and x ∈ B
x ∈ A and x ∈ B
⎡p ⎤
x ∈ A and ( x ∈ B or x ∈ C ) ⎢∴ p ∨ q ⎥
⎣ ⎦
x ∈ A ∩ (B ∪ C)
∴ ( A ∩ B) ⊆ A ∩ ( B ∪ C )
Case 2: x ∈ A ∩ C
Then x ∈ A and x ∈ C
⎡p ⎤
x ∈ A and ( x ∈ C or x ∈ B ) ⎢∴ p ∨
⎣ q ⎥⎦

x ∈ A ∩ (C ∪ B )
∴ x ∈ A ∩ ( B ∪ C ) - by the commutative law
∴ ( A ∪ C) ⊆ A ∩ (B ∪ C )
Therefore ∴ A ∩ ( B ∪ C ) = ( A ∩ B ) ∪ ( A ∩ C )

Although Evans makes two errors in the final lines of case 1 and 2 it is clear that he fully
understands what he is required to show. The fact that he shows in his margin that he
employed disjunctive addition in order to come to the conclusion that
(x ∈ B or x ∈ C ) must surely be an indication that the student understands that the
rules of inference can be applied in such proofs. It is also perhaps indicative that the
student is using the logic as a construct to reason from. If this is indeed the case, then we
would have succeeded in our primary goal i.e. to get students to argue from a logic
perspective.

Only two students (9% of the total number of students) managed to present a complete
argument.

The results indicate that students at this point in time are not entirely comfortable with
proving equality of sets by means of element arguments.
A summary of the argument analysis of the pre-test of set theory is presented in table 21.
 121

Table 21: Argument analysis of set theory  (pre-test)

score No of students score No. of students
0 3   11 1
1 1 12 1
 
2 0 13 0
3 1   14 1
4 2 15 2
5 3 16 1
6 3 17 0
7 1 18 0
8 0 19 0
9 1 20 2
10 0

6.5.1.3 Set theory (post-test)

The following post-test was administered to students:

For all sets A, B and C prove the following:

( A ∪ B) ∪ C = A ∪ ( B ∪ C )

Solution
Suppose x ∈ ( A ∪ B ) ∪ C , then x ∈ A ∪ B or x ∈ C
Case I: x ∈ A ∪ B
Then x ∈ A or x ∈ B , then x ∈ A or x ∈ B or x ∈ C
And so x ∈ A or ( x ∈ B or x ∈ C ) and thus x ∈ A ∪ ( B ∪ C )
Case II: x ∈ C
Then x ∈ C or x ∈ B and so x ∈ C or x ∈ B or x ∈ A
Then x ∈ B or x ∈ C or x ∈ A and so x ∈ A or x ∈ B or x ∈ C
Then x ∈ A or ( x ∈ B or x ∈ C ) and thus x ∈ ( A ∪ B ) ∪ C
So in either case ( A ∪ B ) ∪ C ⊆ A ∪ ( B ∪ C )
Conversely suppose x ∈ A ∪ ( B ∪ C ) , then x ∈ A or x ∈ B ∪ C
Case I: x ∈ A
Then x ∈ A or x ∈ B and also x ∈ A or x ∈ B or x ∈ C
Thus ( x ∈ A or x ∈ B ) or x ∈ C
Hence x ∈ A ∪ ( B ∪ C )
Case II: x ∈ B ∪ C
Then x ∈ B or x ∈ C and also x ∈ B or x ∈ C or x ∈ A
And so x ∈ A or x ∈ B or x ∈ C , then ( x ∈ A or x ∈ B ) or x ∈ C
Thus x ∈ ( A ∪ B ) ∪ C
 122

So in either case A ∪ ( B ∪ C ) ⊆ ( A ∪ B  ) ∪ C
Therefore ( A ∪ B ) ∪ C = A ∪ ( B ∪ C )  
Note: This proof could be done using the associative law. If a student used the associative
law he/she would use fewer steps. In such a  case the marks would then be scaled up to
28.
 
6.5.1.4 Analysis of student answers

22 students completed the post-test. Seven students (32% of the total number of students)
scored 32% or less. The remaining 15 students (68% of the total number of students) all
scored above 53%. Twelve students (55% of the total number of students) had scores of
67% or more. Six students (27% of the total number of students) obtained a score above
80%.
Marsha MacMahon showed major improvement in her ability to prove such statements.
The following is her solution:
Suppose x is a particular but arbitrarily chosen element of ( A ∪ B ) ∪ C , then
x ∈ ( A ∪ B ) ∪ C hence x ∈ A or x ∈ B or x ∈ C
Case I: x ∈ A then
x ∈ A then x ∈ A ∪ ( B ∪ C ) - definition of union
Case II: x ∈ B then
x ∈ B then x ∈ A ∪ ( B ∪ C ) - definition of union
Case III: x ∈ C
x ∈ C then x ∈ A ∪ ( B ∪ C ) - definition of union
Since in all 3 cases x ∈ A ∪ ( B ∪ C ) we have ( A ∪ B ) ∪ C ⊆ A ∪ ( B ∪ C )
If x ∈ A ∪ ( B ∪ C ) then x ∈ A or x ∈ B or x ∈ C
Case I: x ∈ A then
x ∈ A then x ∈ ( A ∪ B ) ∪ C
Case II: x ∈ B
Then x ∈ B then x ∈ ( A ∪ B ) ∪ C
Case III: x ∈ C
Then x ∈ C then x ∈ ( A ∪ B ) ∪ C
Therefore in all 3 cases x ∈ ( A ∪ B ) ∪ C
Thus ( A ∪ B ) ∪ C ⊆ A ∪ ( B ∪ C )
( A ∪ B) ∪ C = A ∪ ( B ∪ C )

Although Marsha’s solution has a number of gaps in it, it seems that she is now aware of
what she is supposed to show i.e. ( A ∪ B ) ∪ C ⊆ A ∪ ( B ∪ C ) and conversely
A ∪ ( B ∪ C ) ⊆ ( A ∪ B ) ∪ C . She also now makes correct use of cases, whereas
previously she gave the impression that she was confused as to the use of cases.
A summary of the argument analysis of the pre-test of set theory is presented in table 22.
 123

Table 22: Argument analysis of set theory  (post-test)

No. of connections No of students No of connections No. of students
0 0   15 1
1 0 16 1
 
2 0 17 1
3 0   18 0
4 1 19 2
5 0 20 2
6 3 21 2
7 0 22 0
8 2 23 1
9 1 24 3
10 0 25 1
11 0 26 0
12 0 27 0
13 0 28 1
14 0

6.5.1.5 Conclusions based on the results of set theory

With the pre-test 65% of students attained a score of less than 45%, whereas with the
post-test only 32% of students had a score less than 45%. The majority of students
therefore obtained a score of more than 53% in the post-test. This is indicative of a major
improvement in proving skills as far as proving equality of sets by means of element
argument is concerned. The three students that have used the rules of inference in their
arguments in the pre-test all scored above 85% in the post-test. It would appear that the
students that used the rules of inference performed better than the majority of students in
both the pre- and post-test, which is perhaps an indication that a thorough understanding
of mathematical logic and its applications can lead to improved deducing abilities.
 124

6.5.2  
Method of direct proof and divisibility

6.5.2.1 Pre-test  
Prove the following by using the method of direct proof:
For all integers a, b and c , if a/b and a/c then  a/(b + c)
 
Solution
Suppose a, b and c are particular, but arbitrarily chosen integers such that a/b and a/c
Then by definition of divide b = ak and c = al for some integers k and l
Then b + c = ak + al = a(k + l)
Let t = k + l, then t is an integer, since the sum of integers is an integer
Thus b + c = at and hence a/(b + c) - by definition of divide

The solution therefore required the students to make the following connections:
1. Use definition of divisibility
2. closure of integers under addition
3. factorize and come to conclusion

6.5.2.2 Analysis of student solutions

22 students attempted the test. Nine students (41% of the total number of students) could
not make any connection and therefore had a zero score. Three students (14% of the total
number of students) could only make one connection. Ten students (45% of the total
number of students) made all three necessary connections.

Marsha MacMahon is an example of a student that had a zero score. Her solution is as
follows:
∀ integers a, b and c, if a/b and a/c then a/(b+c)
Suppose a, b and c are particular but arbitrarily chosen integers such that
If a/b and a/c then a/(b+c)
a a 2a
+ =
b c b+c
2a a
∴ ≠
b+c b+c

∴ the statement is false

Marsha assumes the part that she must prove is true. The fact that she assumes what she
must prove shows that she does not understand the structure of direct proof. She commits
a very basic error when she adds two fractions and adds both the numerators and
denominators. This is an indication of a lower level skill that has not been completely
mastered. She then concludes that the statement is false based on the fact that the two
fractions are not equal.
 125

Siyabonga Maki is an example of a student  that could only make one connection. His
solution is as follows:
 
∀ integers a, b and c, if a/b and a/c then a/(b+c)
  ae
therefore if a/b then b = ad and if a/c then c =
a 1
  =
therefore a/ (b+c), then a/(ad+ae), then a(d + e) d + e
we know that 1/everything
∴ our statement is true.

Siyabonga started off correctly by using the definition of divide, he however then also
used what he is supposed to prove to then deduce an erroneous statement. He then used
the result of this incorrect statement to conclude that the given statement is true. This
shows that he knew he had to use the definition, but did not know how to use this to get
to the desired result. It also shows that he does not quite understand the structure of direct
proof.

6.5.2.3 Post-test

Prove the following by using the method of direct proof:

For all integers a, b and c, if a/b and a/c then a/(b – c)

Solution
∀ a, b and c ∈ Ζ if a / b and a / c then a / b − c
Suppose a, b and c are particular, but arbitrarily chosen integers such that a/b and a/c
Then by definition of divide b = ak and c = al for some integers k and l
Then b – c = ak – al= a(k – l)
Now k – l is an integer since integers are closed under subtraction.
Hence by definition of divide a/ b – c which is what was required.

The solution therefore required the students to make the following connections:
1. definition of divisibility
2. closure of integers under subtraction
3. factorize and come to a conclusion.

6.5.2.4 Analysis of student solutions

22 students attempted the test. One student made 2 connections. The remaining 21
students all gave a complete and correct solution. This implies that 95% of students gave
a complete and correct solution to the post-test. Compare this to the pre-test where only
45% gave a complete and correct solution. This is a dramatic improvement, and I think
indicates that almost all the students have shown some mastery of this kind of problem
since even the student that had one error in his solution showed that he knew what was
required by such a proof. The following is his solution:
 126

Suppose a,b and c are particular but arbitrarily   chosen integers such that a/b and a/c.
Then b = af and c = aj for some integers f and j
Then b- c = af – aj  
b – c = a(f – j)
Now let f – j = l. then l will be an integer and  thus b – c = al.
It is clear that this student is cognizant of the structure of such a proof as all the indicated
 
steps are completely correct. He however neglects to make the last deduction i.e. to
deduce that a/(b – c).

6.5.3 Method of direct proof and number theory

6.5.3.1 Pre-test
Prove the following statement using the direct method of proof:

1
For all integers m, if m > 1, then 0 < <1
m

Solution:
Suppose m is a particular, but arbitrarily chosen integer such that m > 1 , then
1 1
m× 〉 1× (sin ce m >1)
m m
1
hence 1 〉
m
1
now 〉 0, sin ce m > 1
m
1
and hence 0 〈 〈1
m

6.5.3.2 Analysis of student solutions

22 students attempted the test. Fourteen students (64% of the total number of students)
returned a zero score. Five students (23% of the total number of students) had one
connection between statements. Only 3 (14% of the total number of students) students
returned a complete and correct solution. Yasser Buchana who had a zero score presented
the following solution:
1
For all integers m, if m > 1, then 0 < < 1,
m
1
∃ integers m, such that m > 1 and 0 < < 1
m
m > 1 means 1 ≤ m ≤ m + 1
 127

Yasser’s argument is muddled and it is clear  that he is not following the structure
required of this proof. Furthermore he introduces ≤ which shows that he is not fully
 
acquainted with the properties of real numbers.
 
Siphokazi Ncwaiba who made one connection gave the following solution:
 
Lets say m > 1
m 1
÷ 〉
m m
1
1−1 〉 −1
m
1
0〉 −1
m
Siphokazi starts off with the correct reasoning by dividing by m on both sides of the
1 1
inequality to obtain the that is required. She then attempts to show that 〉 0 , but
m m
reasons along an erroneous path and ends up with the opposite of what she intended. She
stops her argument abruptly when she realizes that her argument will not produce the
desired result. Her solution shows however that she is aware of the structure and
requirements of such a proof.

6.5.3.3 Post-test

Prove the following statements using the method of direct proof:

For all real numbers x, if 0 < x < 1 then x 2 < x

Solution
Suppose x is a particular but arbitrarily chosen real number such that 0 < x < 1
multiplying by x we get: 0 < x 2 < x, sin ce x > 0
hence x 2 < x which is what was required.

6.5.3.4 Analysis of student solutions

19 students attempted the test. Eighteen students (95% of the total number of students)
gave a complete and correct solution. One student returned a solution with one error in it.
The results of the pre-test indicate that initially students struggled with this kind of proof,
but the results of the post-test show that after intervention the majority of students are
better equipped to deal with this kind of proof.
 128

6.5.4 Conclusions based on the results of  direct proof

This kind of proof is based on the method  of generalizing from the generic particular.
These proofs are of the form “if P(x) then Q(x”) where P(x) is known as the hypothesis
and Q(x) as the conclusion. To show that “if  P(x) then Q(x)” is true, one supposes that
P(x) is true and then shows that Q(x) must also be true. In other words in order to prove a
 
statement of the form " ∀ x ∈ D, if P ( x) then Q( x) " you suppose that x is a particular
but arbitrarily chosen element of D that satisfies P(x), and then you show that x satisfies
Q(x).
Since there is such a vast number of problems that can be posed where direct proof is
required one cannot conclude, based on the above results, that the experimental group of
students will now be able to solve all such problems. Based on the above argument one
can however make a conclusion that students are now aware of the structure and
requirements of such proofs since the majority of students in both post-tests presented
arguments that contained the correct structure. In other words the majority of students
correctly structured their proofs by supposing x is a particular but arbitrarily chosen
element of the hypothesis and then attempted to show that the conclusion is also true.

6.5.5 Method of ordinary induction (number sequences)

6.5.5.1 Pre-test
Use mathematical induction to prove that:
2 + 4 + 6 + …2n = n2 + n, for all integers n ≥ 1
Solution
Basis step: If n = 1, then
LHS: 2(1) = 2 RHS: 12 + 1 = 2
∴LHS = RHS
∴formula is true for n = 1
Inductive hypothesis: Suppose the formula is true for some integer k ≥ 1 ,
2 + 4 + 6 + …+2k = k2 + k
We must show: 2 + 4 + 6 + … 2(k + 1) = (k + 1)2 + (k + 1)
Inductive step:
Now 2 + 4 + 6 +…+ 2k + 2(k + 1)
= k2 + k + (k + 1)
= k2 + k + 2k + 2
= k2 + 3k + 2
= (k + 1)(k + 2)
Also
(k + 1)2 + k + 1
= k2 + 2k + 1 + k + 1
= k2 + 3k + 2
= (k + 1)(k + 2)
∴
LHS = RHS
This is what was to be shown.
 129

6.5.5.2 Analysis of student solutions  

23 students attempted the test. Four students   (17% of the total number of students)
attained a zero score. Six students (26% of the total number of students) achieved a score
of 20%. Six students (26% of the total number   of students) achieved a score of 40%.
Two students (9% of the total number of students) achieved a score of 60%. Five students
 
(22% of the total number of students) achieved a score of 100%.
All the students that had a zero score could not even start the solution and therefore wrote
nothing. The majority of the students that scored 20% proved the basis step but could not
proceed beyond this. Carmen Williams’ solution is an example of this. The following is
her solution:
LHS: 2(1) = 2
RHS: 12 + 1 = 2
Sk: 2 + 4 + 6 +…2k = k2 + k
SLH: 2 + 4 + 6 + …2(k + 1) = (k + 1)2 + k + 1
= k + 1[(k + 1) + 1]
It seems like she had a general idea of what she is supposed to show, but lacked the
ability to make the necessary logical connections between her steps.
It seems like Siphokazi Ncwaiba understood what the structure of the proof entails and
therefore showed all the requisite steps but could not master the basic algebra that was
required to complete the proof. The following is her solution:
1. Prove that 2 + 4 + 6 + …2n = n2 + n is true for n ≥ 1
LHS: 2n = 2(1) RHS: n2 + n = 12 + 1
= 2 =2
∴LHS = RHS

2. Suppose that it is true for P(k)

2 + 4 + 6 + …2k = k2 + k for k ≥ 1
3. Use P(k) to prove that P(k + 1) is true for k + 1 ≥ 1
2 + 4 + 6 +…2(k+1) = (k+1)2 + (k+1) = k2 +3k + 2
= 2(k+1) = 2k + 2
= (k2 + k) + 2 = k2 + k + 2

6.5.5.3 Post-test ordinary induction (number sequences)

Use mathematical induction to prove that:

1 + 5 + 9 + …(4n – 3) = n(2n – 1) for all integers n ≥ 1

Solution
Basis step: if n = 1, then
LHS: [4(1) – 3] = 1 RHS: 2[2(1) – 1] = 1
∴formula is true for n = 1
Inductive hypothesis:
Suppose the formula is true for some integer k ≥ 1 ,
1 + 5 + 9 + …(4k – 3) = k(2k – 1)
 130

Inductive step:  
We now need to show that it is true for n = k + 1
LHS: 1 + 5 + 9 + … (4k – 3) + [4(k + 1) – 3] 
= k(2k – 1) + 4(k + 1) – 3
= 2k2 – k + 4k + 4 – 3  
2
= 2k + 3k + 1
 
= (2k+1)(k+1)

RHS: (k+1)[2(k+1) – 1]
= (k+1)(2k+1)

∴LHS = RHS
Which is what was required.

6.5.5.4 Analysis of student solutions

22 students attempted the test. Five students scored 50% or less. Two scored 30%, one
scored 40% and two scored 50%.Seventeen students attained a score of 60% or more. Of
these one scored 60%, one scored 70%, one scored 80%, one scored 90% and 13
achieved a 100% score.

Sigqibo Lande was one of the students that had the lowest score. His solution is as
follows:
For n = 1
4(1) – 3 = 1 = 1[2(1) – 1] = 1
∴LHS = RHS
Therefore the statement is true for n = 1
Suppose P(k) is true for some integer k ≥ 1
1 + 5 + 9 + …(4k – 3) = k(2k – 1)
We must show that it is true for k+1. Now if n = k + 1
1 + 5 + 9 + …(4k – 3) + (4k – 2) = k(2k – 1) + k(2k)
∴ k(2k – 1) + (4k – 2)
2k2 – k + 4k – 2
2k2 + 3k – 2
It is clear that Sigqibo understands what he is required to show. However he made some
elementary errors and therefore could not present a concise argument. The majority of the
students that did not succeed in giving a complete and correct solution also made
elementary algebraic errors. This is an indication that students understand what is
required, but struggle with lower level skills. In other words skills students are supposed
to have mastered at school level.
 131

6.5.5.5 Conclusions based on the results of  ordinary induction

With the pre-test 61% of students attained a  score of 40% or less, whereas with the post-
test only 14% of students achieved a score of 40% or less. Furthermore 59% of students
 
achieved a 100% score with the post-test whereas only 22% of students achieved 100% in
the pre-test. This surely then must be an indication that students have acquired the
 
necessary skills as far as this kind of proof is concerned, although there is a strong
indication that lower level skills need serious attention.

6.5.6 Method of strong mathematical induction (recursive sequences)

6.5.6.1 Pre-test
Suppose that c0 , c1 , c 2 , ... is a sequence defined as follows:
c0 = 2, c1 = 4, c2 = 6
ck = 5 ck −3 for all int egers k ≥ 3
Prove that c n is even for all integers n ≥ 0

Solution
Basis step:
Choose a = 0, b = 1, c = 2
c0 = 2 = 2 × 1, which is even
c1 = 4 = 2 × 2 , which is even
c2 = 6 = 2 × 3, which is even
hence it is true for a, b and c

Inductive hypothesis:
Let k be an integer such that k ≥ 3
Suppose P(i) is true for all integers i, with 0 ≤ i ≤ k .
That is c i is even.

Inductive step:
Now since k ≥ 3 it follows that k − 3 ≥ 0, thus 0 ≤ k − 3 < k
then ck − 3 is even since k − 3 < k
hence by definition of even ck − 3 = 2l for some integer l
c k = 5 c k −3
then = 5 (2l )
= 2 (5l )
now 5l is an integer since products of integers are integers, so c k is even.
This is what was to be shown.
 132

6.5.6.2 Analysis of student solutions  

 
22 students attempted the test. Eleven students returned a zero score. Five students
achieved a score of 12,5%. Three students attained a score of 50%. One student had a
score of 62,5% and one student achieved   a score of 75%. Hence 77% of students
achieved a score of 25% or less. Only five students managed to score above 50%.
 
The majority of the students that returned a zero score could not even prove the basis
step. Some could not even start the proof. Others of this group did have some argument,
but the arguments unfortunately were mostly of the muddled variety. For example Yasser
Buchana advanced the following argument:
c3 = 5 c3 − 3 = 5 c0 = 5 (2) = 10
c4 = 5 c 4 −3 = 5 c1 = 5 (4) = 20
c5 = 5 c5 − 3 = 5 c 2 = 5 (6 ) = 30
∴ true for all int egers k ≥ 3

Choose a = 0, b = 1
Suppose ci = 5 ci − 3 is true for integers i ≥ 3
c i + 1 = 5 c ( n + 1) − 3

It seems like Yasser initially tried to check if the sequence holds true for iterations from 3
onwards. He however concludes after only three iterations that the sequence is even for
all integers k ≥ 3 . Subsequently he chose basis values, but did not attempt to prove that
the sequence holds true for these basis values, instead immediately after this he tries to
set up a hypothesis, but then made a nonsensical deduction from this hypothesis. All of
this leaves one with no choice, but to conclude that Yasser does not understand how to
use strong mathematical induction to do the required proof.

Very few students made significant progress in proving the statement; however some did
show that they are aware of the requirements of such proofs. In other words they
attempted to apply the structure of the proof. The following solution of Evans Segodi is
an example of this:

Choose a = 0 and b = 2
Basis step:
P (0) = c0 = 2, which is even
P (1) = c1 = 4, which is even
P (2) = c2 = 6, which is even

Therefore P (0), P (1) and P (2) are true.

 133

Inductive hypothesis: Let k be any integer  k > 2 and suppose ai is even for some
integer i with 0 ≤ i < k  

Inductive step  
We have ck = 5 ck −3 for all integers k ≥ 3
 
k −3≥0
∴ 0≤k −3<k
hence ck − 3 is even
5 ck − 3 is an even integer because it is the multiple of an even integer and so c k is even.
Therefore cn is even for all integers n ≥ 0 .

Evans started off correctly by attempting to prove the basis step. He however does not
choose all of the desired basis values and also fails to show why the initial values hold
true. He proceeds to the inductive hypothesis, but erroneously decides to let k > 2 .
Consequently he proceeded to the inductive step where he correctly showed
0 ≤ k − 3 < k and concluded that c k − 3 is even. He deduces that 5 ck − 3 is even but fails
to prove this by means of the definition of even. Although Evans’ solution can by no
means be described as a complete and rigorous proof it does have the correct structure
and contain most of the elements that are required by such a proof.

6.5.6.3 Post-test

Suppose a1 , a 2 , a3 , ... is a sequence defined as follows:

a1 = 1, a 2 = 3,
a k = a k − 2 + 2a k −1 for all int egers k ≥ 3.
Prove that a n is odd for all integers n ≥ 1

Solution
Basis step:
Choose a = 1 and b = 2, then
a1 = 1 = 2(0) + 1, which is odd by definition of odd
a 2 = 3 = 2(1) + 1, which is odd by definition of odd

Inductive hypothesis:
Let k be an integer such that k ≥ 3
Suppose P(i) is true for all integers i with 1 ≤ i < k , that is ai is odd.

Inductive step:
Now since k ≥ 3 we have k ≥ 2 + 1 , then k − 2 ≥ 1 and k − 1 ≥ 2 ≥ 1,
Hence 1 ≤ k − 2, k − 1 < k and so a k − 2 and a k − 1 are odd, so
 134

a k − 2 = 2l + 1 and a k − 1 = 2m + 1 for some  integers l and m, then

a k = a k − 2 + 2a k − 1  

= 2l + 1 + 2(2m + 1)  
= 2l + 1 + 4m + 2
 
a k = 2 (l + 2m + 1) + 1
Now l + 2m + 1 is an integer, say p and hence a k = 2 p + 1
Thus a k is odd by definition, which is what was to be shown.

6.5.6.4 Analysis of student solutions

23 students attempted the test. One student had a zero score. Thirteen students achieved a
score of 42% or less, with five of these scoring 25%. Ten students achieved a score of
50% or more, with four of these achieving a score of 75% and 3 students attaining a
score of 92%. There was thus a general improvement in results, although the majority of
students still scored less than 50%. Yasser Buchana again had a zero score, whereas
Evans Segodi improved from 62,5% to 75%. The following is Evans’ solution:

Choose a = 1 and b = 2
Basis step: P(1) = 1 which is odd
P(2) = 3 which is odd
Therefore it is true for P(1) and P(2)

Inductive hypothesis: (P(n): an is odd)

Let k be a particular but arbitrarily chosen integer k ≥ 2 and suppose P(i) is true for all
integers i with 1 ≤ i < k , that is ai is odd.

Inductive step:
Given a k = a k − 2 + 2 a k − 1 for all int egers k ≥ 3
Since k ≥ 3 then k − 2 ≥ 1 and k − 1 ≥ 2, hence 1 ≤ k − 2, k − 1 < k
Therefore a k − 2 is odd, a k − 2 = 2l + 1 for some integer l.
And also a k − 1 , is odd, a k − 1 = 2 j + 1 for some integer j
a k = (2l + 1) + 2 (2 j + 1)
= 2l + 1 + 4 j + 2
a k = 2 (l + 2 j + 1) + 1
So
l + 2j + 1 is an integer, therefore ck is odd.
This is what was required to prove.
 135

Although Evans erroneously chooses k ≥ 2   in his inductive hypothesis and he does not
show why the initial values are odd, this solution is more comprehensive than his solution
 
for the pre-test. It is clear therefore that Evans understands what one needs to show in
these kinds of proof.  

6.5.6.5 Conclusions based on the results of  strong mathematical induction

Although there was a general improvement in results, the majority of students still scored
below 50%. This shows that students are still struggling to come to grips with this kind of
proof. However with the pre-test half of the students had a zero score whereas with the
post-test only one returned a zero score. Also only two students attained a score above
50% with the pre-test whereas ten students maintained a score of 50% or higher in the
post-test. So although it seems that the majority of students still did not completely
master this type of proof there was significant improvement in proving ability as far as
adherence to the structure of the proof is concerned. This is corroborated by the fact that
most students showed all of the necessary steps in their post-test.
 136

 
CHAPTER 7

 
STATISTICAL ANALYSIS OF RESULTS

7.1 Introduction  

 
In order to determine if doing a course in logic can improve the mathematical statement
proving abilities of first year mathematics students we compared the pre- and post-test
scores of the logic component to the pre- and post-test scores of the different types of
proofs that we dealt with in our course. As we have previously indicated the pre- and
post-tests of the logic component and the proof component were our main measuring
instruments. Thus in order to do this comparison the various components of the Logic
tests were used to construct an overall logic score that has a possible range of 0 to 100.
This was done separately for the pre- and post-tests. A similar construction was done to
obtain a score for the proofs where again the pre- and post-tests were dealt with
separately. Table 23 shows the terminology and the meanings used in the statistical
analysis:

Table 23: Terminology and meanings used in the statistical analysis

TERM MEANING
Logic pre Sum of all logic pre-tests expressed as %
Logic post Sum of all logic post-tests expressed as %
Proof pre Sum of all proof pre-tests expressed as %
Proof post Sum of all proof post-tests expressed as %
Logic change Sum of logic post minus sum of logic pre
Proof change Sum of proof post minus sum of proof pre

Since our primary interest was in examining the association between logic (as measured
by the tests) and ability in proofs (as measured by those tests) we compared the following
components for the experimental group: logic pre versus proof pre; logic post versus
proof post; logic change versus proof change. Hence the logic scores were the assigned
independent variable and the proof scores the dependent variable. The obvious reason for
this is because by convention the independent variable (x) is assumed to be the one that
causes or explains the variation in the dependent variable (y).
Correlation is a statistical technique used to measure the relationship between two
variables. To determine whether a relationship exists between the variables of logic and
ability in proofs the Pearson product moment correlation coefficient and Spearman rank
correlation coefficient were used. Importantly however, these coefficients also measure
the degree of relationship between the two variables. The values of these correlation
coefficients range from -1.00 to +1.00, with 0.0 indicating no relationship between the
variables, +1.00 indicating a perfect positive relationship and -1.00 indicating a perfect
negative relationship. A positive correlation coefficient indicates that those individuals
who scored high on one variable also tended to score high on the other. A negative
correlation indicates that when the value on one variable is high, it will be low on the
other. The closer the correlation coefficient gets to +1.00 or -1.00 the stronger the
correlation and the closer it gets to 0.00 the weaker it is. The Pearson coefficient is a
 137

 
parametric test whereas the Spearman coefficient is a non-parametric test. Non-
parametric tests are tests where the dependent variables are ranks i.e. where the data are
ranked according to some criteria, whereas  for parametric tests the data need not be
ranked. Spearman is a special case of the Pearson product moment and is most often
  than 20, which was the case for some of
used when the number of pairs of scores is less
our data. The Spearman test utilizes ranked scores and in our case the scores were ranked
 
from low to high. If a positive association exists, we would expect to see a positive
correlation in the logic and proof scores, both pre and post. It is also reasonable to expect
that a positive change in logic scores would be positively correlated with a positive
change in proof scores. In other words if a student improved in logic then that
improvement would result in an improvement in ability in proving mathematical
statements. Besides the fact that we need to determine if there was a relationship between
the variables we also needed to check if the relationship (if there is one) occurred by
chance or not. Measures of statistical significance tell us the probability that the
association occurred by chance. We will check for statistical significance at the .05 level.
Significance at the .05 level means that only 5 times out of 100 the results obtained
occurred by chance alone, therefore the probability that it occurred by chance is at the 5%
level. Hence any probability less than .05 will be accepted as significant.

The null hypothesis for both Pearson and Spearman is ρ = 0. In other words the null
hypothesis is that there is no relationship between the two variables.
The low number of observations can be ascribed to the fact that not all students
completed the tests. Only those students that had all the relevant data could be taken into
consideration. Since the number of pairs of scores was lower than 20 the Spearman test is
the preferred test.

7.2 Logic pre versus Proof pre

The Pearson correlation coefficient for logic pre versus proof pre is 0.35768 and for
Spearman it is 0.31098 indicating in both cases a weak positive relationship. Both
correlation coefficients are also not statistically significant at the .05 level since both have
probabilities much greater than .05. Table 24 contains the result for logic pre versus
proof pre:

Table 24: Comparison of Logic pre to Proof pre

Pearson correlation coefficient 0.35768
Probability of significance 0.3102
Number of observations 10
Spearman correlation coefficient 0.31098
Probability of significance 0.3818
Number of observations 10
 138

7.3 Logic post versus Proof post  

For logic post versus proof post the Pearson   coefficient is 0.18250 and for Spearman it
is 0.21366 which indicates a very weak positive relationship. The correlation coefficients
are not significant since it is greater than .05.  The result for logic post versus proof post is
shown in Table 25:
 
Table 25: Comparison of Logic post to Proof post.
Pearson correlation coefficient 0.18250
Probability of significance 0.5702
Number of observations 12
Spearman correlation coefficient 0.21366
Probability of significance 0.5049
Number of observations 12

7.4 Logic change versus Proof change

For logic change versus proof change the Pearson coefficient is 0.42407 and the
Spearman is 0.35000 which also indicates a weak relationship. The correlation
coefficients are not statistically significant at the .05 level since both have probabilities
much greater than .05. The result for logic change versus proof change is shown in
table 26:

Table 26: Comparison of Logic change to Proof change

Pearson correlation coefficient 0.42407
Probability of significance 0.2553
Number of observations 9
Spearman correlation coefficient 0.3500
Probability of significance 0.3558
Number of observations 9
 139

7.5 Graphs  

A scatter diagram can be used to determine   if an association exists between two

variables. The amount of scatter in a scatter diagram gives us a rough measure of the
 
strength of a correlation and normally the diagram is done first and then the correlation
coefficient is determined to either deny or confirm the findings of the scatter diagram.
 
7.5.1 Logic pre versus Proof pre

The scatterplot of logic pre versus proof pre is consistent with the correlation coefficient
i.e. it shows a weak positive linear relationship. However if points A and B could be
ignored then the amount of scatter would be much less and hence the degree of
association much more. The scatterplot for logic pre versus proof pre is shown in
figure 12.

Figure 12
 140

7.5.2 Logic post versus Proof post  

The scatter diagram of logic post versus  proof post seems to indicate a non-linear
relationship. However all scores for both logic and proof are above 60% whereas the
majority of scores for logic pre and proof pre  were below 50%. The scatter plot for logic
post versus proof post is shown in Figure 13.
 
Figure 13
 141

7.5.3 Logic change versus Proof change  

 
The scatter diagram of logic change versus proof change indicates a weak positive linear
relationship which confirms the correlation coefficient. The distribution of the scatter plot
indicates that there is an even split between  those with a positive correlation between
logic difference and proof difference and those with a negative correlation. The
 
scatterplot therefore indicates that there is no conclusive evidence for a positive
association since it exhibits a weak positive linear relationship. The scatterplot for logic
change vs proof change is shown in Figure 14.

Figure 14

The fact that two variables are correlated does not imply causality; conversely if two
variables are not correlated one cannot be the cause of the other. Since in all three cases
above a weak relationship was indicated there is thus no clear indication that logic had an
effect on proving ability. So, despite the fact that there is indication of a weak positive
relationship in at least two of the coefficients the null hypothesis has to be accepted.
 142

7.6 Control Group versus Experimental  Group

 
We also compared the scores of the experimental and control groups in terms of the first
logic puzzle (pre-test) and the third logic puzzle (post-test). Our first comparison was at
the baseline (pre-test) where the expectation  was that the two groups would be similar.
On this component the scores ranged from 0 to 3. The results of this comparison are
 
shown in Table 27 below:

Table 27: Comparison of Pre-test scores of Control and Experimental groups

PUZZLE I (pre-test)
Control group
Outcome variable 0 1 2 3 Total
Frequency 12 5 3 1 21
Percent 31.58 13.16 7.89 2.63 55.26
Row percent 57.14 23.81 14.29 4.76 100
Experimental group
Frequency 7 7 1 2 17
Percent 18.42 18.42 2.63 5.26 44.74
Row percent 41.18 41.18 5.88 11.76 100
Total 19 12 4 3 38
Total percent 50.00 31.58 10.53 7.89 100.00
Frequency missing = 5

The statistical results based on Table 27 are shown in table 28 below:

Table 28: Statistical results based on table 27

Statistics for Puzzle I
Statistic Degrees of freedom Value Probability
Chi-square 3 2.5901 0.4592
Likelihood ratio chi-square 3 2.6308 0.4521
Mantel-Haenszel chi-square 1 0.4920 0.4830
Phi coefficient 0.2611
Contingent 0.2526
Cramer’s V 0.2611
Fisher’s Exact test: Table probability 0.0166
Probability 0.5109

The following is an explanation of the values in Table 27. The outcome variable is the
score obtained by the students. The frequency indicates the number of students that
obtained a score. For example 12 of the control group and 7 the experimental group of
students scored 0 for the first puzzle. The 21 at the end of the frequency row is the total
number of control group students that participated in the study and the 17 in the
frequency row of the experimental group is the total number of students for the
experimental group. The 38 indicates the total number of students in both the control and
experimental groups.
 143

The percent row indicates the percentage of  students that attained a specific score. For
example the 31.58% of the control group shows that 12 out of 38 (expressed as a percent)
students scored a 0 for puzzle 1.  
The row percent indicates the number of students of that specific group that obtained a
 
specific score. For example 7 out of 17 students (41.18%) of the experimental group
scored a zero on puzzle 1.
 
Since more than 50% of the cells have expected counts less than 5 the chi-square statistic
might not be valid and therefore Fisher’s exact test is preferred. If the probability of the
Fisher’s exact test is less than 0.05 then there will be a significant difference between the
puzzle scores of the control and experimental groups. Since p = 0.5109 for the Fisher test,
the groups do not differ significantly. Therefore at the beginning of the study the abilities
(to solve logic puzzles) of the control and experimental groups of students were similar.

As already indicated our second comparison is on the post-test scores to see if the
experimental group performed better as a result of having been taught the content in the
logic course. The results of this comparison are shown in Table 29:

Table 29: Comparison of Post-test scores of Control and Experimental groups

PUZZLE III (post-test)
Control group
Outcome variable 0 1 2 3 Total
Frequency 5 7 7 2 21
Percent 11.90 16.67 16.67 4.76 50.00
Row percent 23.81 33.33 33.33 9.52 100
Experimental group
Frequency 1 1 4 15 21
Percent 2.38 2.38 9.52 35.71 50.00
Row percent 4.76 4.76 19.05 71.43 100
Total 6 8 11 17 42
Total percent 14.29 19.05 26.19 40.48 100.00
Frequency missing = 1
The statistical results based on Table 29 are shown in Table 30 below:

Table 30: Statistical results based on Table 29

Statistics for Puzzle I
Statistic Degrees of freedom Value Probability
Chi-square 3 17.9260 0.0005
Likelihood ratio chi-square 3 20.0535 0.0002
Mantel-Haenszel chi-square 1 14.5871 0.0001
Phi coefficient 06533
Contingent 0.5469
Cramer’s V 0.6533
Fisher’s Exact test : Table probability 4.002 x 10 -6
Probability 1.524 x 10 -4
 144

The row percent indicates that the percentage   of control group students that achieved a
perfect score have increased from 4.76% to 9.52%, whereas 71.43% of the experimental
  intervention.
group scored full marks for the puzzle after the
Since 50% of the cells have expected counts less than 5 the chi-square statistic may not
 
be a valid test, hence Fisher’s exact test is preferred. The p = 1.524 x 10 -4 for Fisher’s
test which is substantially less than 0.05, hence the experimental and control groups
 
differ highly significantly.

7.7 Stratified Analysis

Using Cochran-Mantel-Haenszel methodology a more meaningful second comparison

was done on the post-test scores. This second comparison utilized a stratified analysis
using the pre-test scores as strata. What this in essence means is that the pre-test scores of
both the control and experimental group of students are compared to their post-test
scores. For example all the students that scored 0 in the pre-test are checked to see what
they scored in the post-test i.e. did they score the same, less or was there an improvement.
The following four tables will investigate this.

7.7.1 Controlling for a pre-score of 0

Table 31 shown below reports on students that scored 0 in the pre-test. This 0 score is
then compared with the post-test i.e. how did these students perform in the post-test. The
following is an explanation of the values in table 8. The outcome variable is the new
score obtained by the students. Four of the 6 students (i.e. 66.67% of students) of the
experimental group that scored 0 in the pre-test scored full marks in the post-test. Only
one of the six (16.67%) again scored 0. The same trend is not seen in the control group
where zero students scored full marks and 41.67% of students again scored 0 and 50%
had improved to a score of 1.

Table 31: Controlling for a pre-score of 0

Control group
Outcome variable (score) 0 1 2 3 Total
Frequency 5 6 1 0 12
Row percent 41.67 50.00 8.33 0.00
Experimental group
Frequency 1 0 1 4 6
Row percent 16.67 0.00 16.67 66.67
Total 6 6 2 4 18

7.7.2 Controlling for a pre-score of 1

Table 32 shown below reports on students that scored 1 in the pre-test. Three of the five
control group students (i.e. 60% of the students) that scored 1 in the pre-test scored 2 in
the post-test and 40% of these students now scored full marks. The experimental group
shows a dramatic improvement as 71.43% of these students scored full marks in the post-
test.
 145

Table 32: Controlling for a pre-score of 1  

Control group
Outcome variable 0 1   2 3 Total
Frequency 0 0 3 2 5
 
Row percent 0.00 0.00 60.00 40.00
Experimental group  
Frequency 0 0 2 5 7
Row percent 0.00 0.00 28.57 71.43
Total 0 0 5 7 12

7.7.3 Controlling for a pre-score of 2

Table 33 shown reports on students that scored 2 in the pre-test. The control group of
students did not show an improvement, in fact one of these students scored one in the
post-test which is worse than the two scored in the pre-test. The remainder of these
students scored a two again and hence did not improve. In contrast the experimental
group student that scored two improved to three in the post-test.

Table 33: Controlling for a pre-score of 2

Control group
Outcome variable 0 1 2 3 Total
Frequency 0 1 2 0 3
Row percent 0.00 33.33 66.67 0.00
Experimental group
Frequency 0 0 0 1 1
Row percent 0.00 0.00 0.00 100.00
Total 0 1 2 1 4

7.7.4 Controlling for a pre-score of 3

Table 34 shown below reports on students that scored 3 in the pre-test. The control group
student that had a perfect score in the pre-test regressed to 2 in the post-test. The two
experimental group students maintained their perfect score.

Table 34: Controlling for a pre-score of 3

Control group
Outcome variable 0 1 2 3 Total
Frequency 0 0 1 0 1
Row percent 0.00 0.00 100.00 0.00
Experimental group
Frequency 0 0 0 2 2
Row percent 0.00 0.00 0.00 100.00
Total 0 0 1 2 3
 146

7.7.5 Cochran-Mantel-Haenszel statistics 

  stratified analysis. Table 35 reports on the

Table 35 is a summary of the statistics for the
Cochran-Mantel-Haenszel test for Puzzle III controlling for Puzzle I. The Row Mean
 
Scores differ significantly at a 1% level of significance (p = 0.0013 therefore p< 0.01).
The significant p-value indicates that the scores obtained by the experimental and control
 
groups are conditionally dependent. It could be seen that the experimental group
improved their prior scores whereas this was not the case for the control group.

Table 35: Cochran-Mantel-Haenszel Statistics (Modified Ridit Scores)

Statistic Alternative Hypothesis DF Value Probability (p)
1 Nonzero Correlation 1 10.0135 0.0016
2 Row Mean Scores Differ 1 10.3371 0.0013
3 General Association 3 15.6618 0.0013
Frequency missing = 6

7.8 Conclusions based on the comparison between control and experimental

groups

The statistical results for the comparison show that the initial deductive abilities of
control and experimental groups were more or less similar. Subsequently the
experimental group was exposed to instruction in a course on logic, whereas the control
group received instruction in a course on Differential Calculus. Approximately 5% of
control group students and 12% of the experimental group attained a perfect score in the
pre-test. With the post-test these figures have increased to approximately 10% and 71%
for the control and experimental groups, respectively. So the number of control group
students that achieved a perfect score had doubled whereas the number of experimental
students had increased six-fold. The stratified analysis also showed that the experimental
group improved consistently, whereas the control groups were inconsistent and did not
show the same dramatic improvement as the experimental group. It is obvious that the
remarkable improvement of the experimental group as far as the logic puzzles are
concerned can only be ascribed to the instruction in logic.

7.9 Summary of conclusions on statistical analysis

The statistical analysis reveals that there is no clear indication that knowledge of logic
ameliorates mathematical statement proving ability. The above discussion however
makes it clear that the experimental group performed better than the control group in the
logic post-test as a consequence of instruction in logic. Since the study did not compare
control and experimental groups with respect to proving mathematical statements, we do
not know if knowledge of logic would have caused the experimental group to perform
better than the control group in this aspect. As a result we can only conclude that
knowledge of logic does contribute to students making more deductions between and
from statements of logic puzzles, than students that have no knowledge of logic.
 147

 
CHAPTER 8

 
SUGGESTIONS, RECOMMENDATIONS, CONCLUSION and FUTURE
RESEARCH
 
8.1 Introduction
 
The International Commission on Mathematical Instruction (ICMI) [48] in their
discussion document on the role of proof and proving in mathematics education states the
following: “The recent NCTM35 Principles and Standards document has elevated the
status of proof in school mathematics, as have several European educational jurisdictions
responsible for the school mathematics curriculum (in the UK, Italy, Spain and
elsewhere). Some developing countries like South Africa also specifically mention the
teaching of other functions of proof such as explanation and the importance of making
and proving or disproving conjectures.” Stylianides and Stylianides [80] shares these
sentiments as can be seen from the following quote: “There are currently increased
efforts to make proof central to school mathematics throughout the grades.” I think it is
clear from the above quotes and the earlier discussion on the importance of proof (pg 21)
that proof and proving is essential in mathematics. However despite the importance of
proof most students of mathematics are not adequately prepared by the school system for
proof and proving at tertiary institutions.

8.2 Phases of Proof

A number of researchers are of the opinion that students are ill-prepared for the rigours of
proving at tertiary level by the proving methods presented to them at school. One of the
reasons advanced for this state of affairs is that students are only presented with the
phases of exactification and application of proof. In other words it has become customary
to teach proof in mathematics by deductively presenting the proof and then asking
students to learn it and then use it to solve homework and examination problems. Some
like Kleiner [52] believe that this deductive presentation might lead students to develop
feelings of inferiority since this might lead students to believe that proofs are created by
geniuses who start with axioms and reason directly and flawlessly to theorems. In other
words students do not associate struggling and failure with proving. Kort et al [55]
therefore argue that teachers fail to teach students that feelings associated with various
levels of failure are normal parts of learning and that these feelings can actually be
helpful signals for how to learn better.

The other phase (the phase of experimentation) involved in proving was not previously
shown to students at school level in South Africa. The National Curriculum Statement
(NCS)36 makes an attempt to change this state of affairs by proposing an investigative
approach to proving. In this approach the phase of experimentation is represented by the
investigation. From the results of the investigation learners are expected to make

35
NCTM – National Council of Teachers of Mathematics.
36
The new South African mathematics curriculum for grades 10 to 12.
 148

conjectures and generalisations. Subsequently   learners are expected to provide

explanations and justifications and attempt to prove conjectures.
 
Students only start developing deductive reasoning skills on the third of the five Van
Hiele levels of reasoning. This I believe  is an indication of the difficulty level of
deductive reasoning.The Van Hieles attributed the dismal performance by students in
 
geometry to the fact that the curriculum was presented at a higher level than those of the
students. As a result the students could not understand the teacher and the teacher could
not understand why the students could not understand. Is this perhaps also the case with
other forms of proof i.e. the fact that proofs are presented to students initially on too high
an abstract level and hence they cannot understand? We are of the opinion that this is the
case hence what is needed is to start proofs from the phase of experimentation so that
students can start the proof with their pragmatic reasoning schemas as proposed by
Cheng et al [14] and agreed to by Stylianides and Stylianides [80]. If this methodology is
followed students will not be required to go straight to deductive reasoning, but will start
with inductive and other forms of reasoning where they can use their pragmatic reasoning
schemas that form part of their natural reasoning abilities. This therefore will allow
students to start with less abstract and deductive skills which are less demanding and
hence allow students to be phased into proving initially using skills they are more
competent in. This methodology might not always be practically possible because of time
constraints. We however strongly advocate that this is the methodology that should be
followed when students are introduced to proof and proving for the first time, since in
this way students will not only understand the need for the proof but would have seen the
development of the proof through all the different phases and in this way perhaps
perceiving that the need for proof is based in practical considerations. The question is
does the investigations prescribed by the NCS mentioned earlier fulfill this role?

My experience has been that the majority of mathematics teachers and advisors in South
Africa are not aware of the different phases of proving and the need for it. They therefore
do not understand the need for investigations and adhere to it only since it is a prescribed
assessment method. The result is that the deductive presentation of proofs is the preferred
pedagogical method and the investigation is done in a very perfunctory manner. The
consequence of this deductive presentation is that the majority of proofs are presented to
learners on an abstract level that is beyond most of them since none of their studies prior
to this has adequately prepared them for the rigours of deductive reasoning. Also our
research has shown that the majority of teachers are not aware of the different types of
proof and do not even know what types of proof are part of the school curriculum (i.e. the
proofs they are teaching). This state of affairs should be addressed as a matter of urgency
especially seen in the light of the renewed emphasis on proof and proving. A possible
solution for this problem is to expose teachers to instruction in proof that include all the
different types of proof, structure of proofs, phases of proof, deductive reasoning,
elementary logic and the development of abstract reasoning in learners. I am of the view
that even the contested views as to what proof is should be included in such instruction.
This instruction should be for both prospective teachers and most importantly also for
teachers already part of the educational system.
 149

8.3 Logic and Deductive reasoning  

Although our research has not conclusively  shown that knowledge of logic improve the
proving abilities of students it is clear that logic does form the basis of the majority of the
proofs that we dealt with. One way of describing   deductive reasoning according to
Stylianides and Stylianides [80] is as the general form of reasoning associated with
 
logically necessary inferences based on given sets of premises. They regard the logical
rules of inference modus ponens and modus tollens as two specific forms of deductive
reasoning. According to them modus ponens forms the foundation of direct proof and the
proof method by mathematical induction, whereas modus tollens is the foundation of
indirect proof which includes proof by contradiction and contraposition. Epp [34] is of
similar opinion as can be seen from the following quote: “It is clear that proof and
deductive reasoning are very closely associated. Deductive reasoning in turn cannot be
separated from logic.” The above and some of our earlier arguments show that logic,
deductive reasoning and proof are intrinsically linked. It is imperative therefore that
teaching and learning of proof should include instruction in logic.

We are of the opinion that learners at high school level should at least be introduced to
the rules of inference since they utilize the rules of inference in the majority of their
proving exercises. For example the type of proof that is prevalent in the high school
system in South Africa is direct proof. In this proof method one argues from a hypothesis
and make deductions until you reach a conclusion. This proof method therefore relies
heavily upon deductive reasoning which in turn is mostly based on the rules of inference.
Furthermore the following quote from Stylianides et al [80] shows how strong the
relationship between deductive reasoning and proof is: “In addition, from a psychological
standpoint the development of students’ ability for deductive reasoning has been found to
go along with the development of their ability for proof.” These I firmly believe are the
most compelling reasons why the rules of inference should be introduced as a topic of
study at school level. We believe that this can be done less formally by utilizing logic
puzzles. The logic puzzles presented to students in our study contained the following
rules of inference: modus ponens, modus tollens and disjunctive syllogism. Hence
puzzles like these can be utilized as an informal introduction to deductive reasoning to
show students how in proving to make deductions using these rules of inference.

Since reasoning by deduction plays such a crucial role in proving I think it is important
that much more research is done in this area utilizing the findings of neuroscience and
cognitive psychology so as to gain a deeper and better understanding of this type of
reasoning. Studies done by Cheng et al [14] has shown that deductive reasoning does not
happen naturally so we need to research and find teaching methodologies that will lead to
the enhancement and development of this kind of reasoning by students. The results of
our study have shown how students have improved in their ability to make connections
(deductions) between statements and to draw conclusions from given hypothesis in logic
puzzles as a result of exposure to the rules of inference. The skills of making connections
between statements, drawing conclusions and reasoning one-step at a time is vital skills
in proof and proving. We therefore emphasize again that Lewis-Carol type logic puzzles
can be utilized as an informal way of introducing students to deductive reasoning and as a
 150

  of logical connectives such as and, or and

tool to make students aware of the implications
not in mathematical statements. We must therefore pay much more attention to the
reasoning, psychological and neural processes  that are involved in deductive reasoning to
improve proof and proving at all levels of education.
 
8.4 Logic and cues
 
Cheng et al [14] are convinced that instruction in logic should be accompanied by
concrete examples to show where and when the logic can be applied in proofs as can be
seen from the following quote: “ …if logic instructors wish to influence their students’
inferential behaviour in the face of novel problems, they must do much more than they
currently do to show how to apply logical rules to concrete problems.” This therefore
mean that if students are instructed in logic without showing them how to use the logic in
concrete examples of proving then the training will not be beneficial. This is perhaps why
our hypothesis was not confirmed since we did not consistently show where the logic can
be applied in specific proving situations. We therefore suggest that the teaching of logic
principles be accompanied by examples in proof where the logic is applied. The opposite
should also be done i.e. whenever proof is taught then cues should be given as to where
and when it is possible to apply the logic principles.

8.5 A Hierarchical and Sequential System for proof

Proof should be done right through the grades as one of the concerns of a number of
researchers has been that students are abruptly introduced to proof in high school and that
this is the reason why students struggle so much with proof. Proof and proving therefore
should form an integral part of all mathematics starting from the lower grades and
continuing through all levels of school and tertiary institutions. It is important that
instructors in proof should be cognizant of the different levels of reasoning involved in
proof. Proof curricula should therefore be devised in such a way so as to allow students to
go through the different levels of reasoning in a hierarchical and sequential manner. Care
should be taken that the level of abstractness of a proof matches the level at which the
student can reason abstractly. Proof instructors should therefore based on these reasoning
levels devise proof curricula that emphasize a hierarchical and sequential approach that
takes into consideration student abstract reasoning development.

The South African revised national curriculum statement (RNCS) for grades R to 9
includes some of the building blocks of the reasoning and skills involved in proving such
as: conjecturing, inferring, deducing, justifying refuting, forming conclusions, etc. This
shows that the curriculum writers attempted to design the curriculum in such a way as to
allow the learners to engage with proof or aspects of proof in a coherent way right
through the grades. Our contention however is that these proof skills are not learned since
learners are struggling to master the basic numeracy skills (as indicated earlier) without
which they cannot proceed to more abstract reasoning skills and as a result the efficacy of
the design is negated.
 151

8.6 Proof structure and practice  

  an important role in students’ ability to do

Stylianides et al [80] argue that practice plays
deductive reasoning and thus in their ability for proof. They state the following in this
regard: “A possible use of practice is to help  students develop the strategies for effective
management of their working memory capacity. In addition, we hypothesize that practice
 
can be used to help students internalize the general logical structure of different proof
methods, such as proof by contradiction, thus releasing working memory capacity to be
spent in the application of these proof methods.” Since I fully concur with this argument
of Stylianides et al I will use an example of proof by contradiction as an explication.

The logical structure of proof by contradiction is as follows:

Step 1: Suppose the statement to be proved is false (i.e. use negation)
Step 2: Show that this supposition leads logically to a contradiction
Step 3: Conclude that the statement to be proved is true.

Example
Prove the following statement by contradiction:
For all integers a, b and c, if a ∤ bc then a ∤ b

Solution

Step 1
Suppose ∃ integers a, b and c such that a ∤ bc and a/b

Step 2
Since a/b there exists an integer k such that b = ak - by definition of divide
Then bc = (ak ).c
bc = a (k .c) - by associative law
But kc is an integer since it is a product of integers and so a/bc - by definition of divide
Thus a ∤ bc and a/bc which is a contradiction

Step 3
Hence the supposition is false and the statement is true.

8.6.1 Discussion

What Stylianides et al [80] is suggesting is that the student needs to make the logical
structure of proof by contradiction (as set out above) part of his/ her long term memory.
In order to do this example the student also needs to have internalized the definition of
divide and the associative law. The working memory can then access this information
from the long term memory to make the indicated connections. The following delineates
how this happens in practice:
- First step one has to be utilized to write the negation. This implies that the student at
an earlier stage has engaged with negations and now has a complete mental schema
 152

of negation. This means that the student   knows that negation in this case means
accepting that the hypothesis is true i.e. a ∤ bc and the conclusion is false i.e. a/b
- Second the student should know from prior   experience that he/ she next has to start
with a/b since the definition of divides is known. This definition should then be
 
applied.
- Next the equation that is a result of the   application of the definition should be
connected to the hypothesis i.e. by multiplying by c on both sides of the equation
giving bc = (ak).c Again the student can only know this as a result of prior
experience that is brought about by practice.
- The student should then realize that the definition of divide can be applied again to
the equation bc = (ak).c but this time in the opposite direction. In other words if the
associative law is applied to give bc = a.(kc), then from this equation one can deduce
that a/bc. This deduction can only be done by comparing the equation to the
definition of divide that is stored in the long term memory.
- The fact that a ∤ bc and a/bc brings about the contradiction that was required.
- Step 3 can now be applied to conclude that the supposition is false and hence the
original statement is true.

It is clear from the above outline that each step in the reasoning process is dependent on
the previous step.

Now if the student has not internalized the structure of the proof then he/ she will have to
keep in the working memory the required definitions and the structure of the proof while
at the same time devising a strategy to induce a contradiction. This obviously limits
working memory capacity and is a much more difficult exercise and we therefore suggest
that the student internalize the structure of the proof by means of practicing on examples
that require few deductions. As the student becomes more confident in applying the
structure of the proof progressively more difficult examples can be provided. This should
be the case for the other types of proof too.

It is clear therefore that students should practice the different proof methods to gain
experience to become confident and importantly to enhance the capacity of their working
memories. The only way in which students can engage successfully with proving then is
to practice so that the proof structure can become part of their long term memory. Their
working memory is then freed for other important functions of proving like looking for
patterns in the given information. Furthermore neuro-science has shown the more you
practice the more permanent connections are formed in the brain and thus long term
memories are created.

The current South African NCS mathematics curriculum specifies that learners are to
engage with proving activities in various content areas like number patterns and
Euclidean geometry. The curriculum however does not make specific mention that
students are to know what proof and proving mean and that there are different methods of
proving and that each proof has a unique structure. The curriculum also does not
prescribe that learners should be made aware of the type of proof they are employing in
their proof activities. We however are of the opinion that learners should be taught which
 153

type of proof they are employing in all their  proof activities. Furthermore learners should
be made aware that each proof has a specific structure and that it is of utmost importance
that they commit the structure of the proof to  their long term memories. This is so that if
with the passing of time they forget the specific proof examples they have dealt with they
will still remember the structure of the proof   and the methodology employed. More
advanced levels of proving at school level should also include instruction on identifying
 
conditions that will indicate that a specific proof cannot be applied.

8.7 Assessment and intervention strategies

The NCS prescribes four types of assessment namely baseline, diagnostic, formative and
summative assessment. Baseline assessment is used to establish what skills and
knowledge learners already have and is usually done at the start of a grade or learning
cycle. Diagnostic assessment is used to uncover the cause or causes of a learning barrier
and therefore assists educators to decide on support strategies. Alternatively it is used to
discover what learning has not taken place so as to put intervention strategies in place.
The goal of formative assessment is to provide feedback to the learner and to inform the
teacher as to the progress of the learner. Summative assessment is used to judge the
competence of learners and is therefore used for progression purposes.

One of the factors that contribute enormously to student deficiency in mathematics and
especially in proof is the fact that although reasoning errors, misconceptions, gaps in
prior knowledge, etc. are uncovered by assessment it is not addressed in the majority of
cases. The following quote from Kutzler [58] provides a reason for this state of affairs:
“The curriculum forces the teacher to continue with the next topic independent of the
progress of individual students.” What Kutzler is alluding to here is that teachers are
forced by time constraints, volume of the content of the curriculum and other factors like
class size to continue with the next topic although some learners might not have acquired
the requisite competency in the previous topic. As indicated earlier what this then implies
is that learners are passed on from grade to grade without acquiring the necessary
competencies. This in turn will prevent students from building a complete understanding
of mathematical concepts. The question is what can be done to alleviate these
deficiencies?

How does one know if a student understood a mathematical lesson that was presented in
class? In general student reasoning can be determined either by verbal responses or by
written responses in answer to assessment. Hence teachers have to utilize either verbal
questioning or paper and pencil assessment to determine students’ levels of
comprehension. Teacher questioning strategy whilst presenting lessons and also on
completion of lessons is of utmost importance and should be based on the cognitive
requirements of the topic under discussion. One of the aims of a questioning strategy
obviously has to be to determine if the students understood the teaching, but also and
most importantly to determine student misconceptions and gaps in prior knowledge. Once
student errors in reasoning, misconceptions or gaps in prior knowledge are determined
these have to be dealt with immediately so as to eliminate these problems, but also to
prevent students progressing to the next grade with these misconceptions and incomplete
 154

understanding of mathematical concepts. So  in essence remedial teaching should be an

ongoing teaching strategy. It would therefore be in the interest of students and teachers to
spend more time on baseline, diagnostic and   formative assessment than on summative
assessment. We are aware however that large class sizes and the administrative load of
the average teacher in South Africa are   inhibiting factors that prevent this from
happening on an ongoing basis.
 
8.8 Mathematical language versus Everyday language

There is a definite difference between everyday language and mathematical language.

Epp [34] argues about this as follows: “ One reason students may have problems with
formal mathematical reasoning is that certain forms of statements are open to different
interpretations in informal and formal settings…By contrast, mathematical language is
required to be unambiguous, with each grammatical construct having exactly one
meaning.” This difference between everyday and mathematical language complicates
proof and proving even more since every conjecture and statement that has to be proved
is given in precise mathematical language. Failure to interpret the statement correctly
leads to all kinds of complications in proving. It is therefore incumbent upon proof
instructors to teach students how to interpret mathematical statements. The results of our
study have shown that instruction in the logic of compound and quantified statements
provides an effective way to achieve this.

8.9 Pedagogical content knowledge

We have indicated in some of our earlier arguments that the main protagonists in the
teaching and learning process are the teacher and the learner. Note we are not arguing
that they are the only role players, but that they are the main ones. As a result of this it is
imperative that the pedagogical content knowledge of teachers be improved. Pedagogical
content knowledge refers to the knowledge that is required by teachers to know how to
teach a specific concept in mathematics. Research done by both provincial and national
education departments in South Africa indicate that the mathematical results need a lot of
improvement. One of the reasons for the bad results in our opinion is that teachers
currently in the system are not equipped pedagogically for the demands of the new
curriculum both at the primary and high school level. This is corroborated by the
TIMSS [47] report that showed that South African teachers are among the lowest
qualified mathematics teachers of the 50 countries that participated in the study. What we
are advocating is that much more attention should be given to those pedagogical skills
that are required by the new curriculum. This cannot be accomplished by short once off
workshops as is the case currently. Teachers in the system should be allowed study leave
so that they can for extended periods study and research the pedagogical content
knowledge that is required by the new curriculum. Alternatively in service training
should be provided on an ongoing basis and not on a once off basis as is currently the
case.

8.10 Conclusion
 155

 
It is clear that high school learners leave the school system with less than the expected
competencies in terms of proof and proving.   There can be no doubt that proof and
proving is essential in mathematics. It has been shown that the cognitive abilities
developed by reasoning in proof are abilities  that can be employed in other domains of
mathematics and also in other subject areas. Some of our earlier arguments have also
 
indicated that the development of good science and mathematics teachers is crucial to the
economic development of countries. All of these arguments therefore compel learning
institutions to spend more time and effort to develop the necessary competencies in
students in terms of proof and proving. The arguments in this paper therefore have to be
seen as an attempt to provide some solutions and debate around the essential topic of
proof and proving. Finally I would like to agree with Stylianides and Stylianides [80] that
findings of different disciplines need to be harnessed to truly understand the learning
process as far as proof and proving are concerned. We therefore suggest that in order to
truly understand how learners reason in proof specifically and mathematics in general and
hence to correct reasoning errors and to teach better one has to cross the divide between
mathematics education, educational psychology and neuroscience.

8.11 Future Research

Based on the above arguments I am of the opinion that an area of research that requires
our attention is the reasoning involved in the different kinds of activities in the
mathematics classroom. Our suggestion is that the different kinds of reasoning should be
investigated in the light of current findings of neuroscience, educational psychology and
teaching methodologies so as to deliver a better pedagogical product, to improve the
learning process in mathematics to ultimately make mathematics more understandable
and enjoyable for both teacher and learner.

References
 156

 
1. Anderson, J.R., Reder, L.M., & Simon, H.A. (2000, Summer). Applications and
 
Misapplications of Cognitive Psychology to Mathematics Education. Texas
Education Review.
2.  
Ayalon, M., Even, R. (Feb. 2007). Mathematics Learning and the Development
of General Deductive Reasoning. Proceedings of the Fifth Congress of the
 
European Society for Research in Mathematics Education. Larnaca, Cyprus:
University of Cyprus
3. Barker, S.F. (1964). Philosophy of Mathematics. New Jersey: Prentice Hall inc.
4. Berger, M. (June 2006). Making mathematical meaning: from preconcepts to
pseudoconcepts to concepts. Pythagoras, 63, 14 – 21.
5. Berriozabal, M. (1995). The Texas prefreshman engineering program: Filling the
pipeline for workforce diversity. Proceedings of the National Symposium and
Career Fair of the Society of Mexican-American Engineers and Scientists.
6. Blanton, M. , Stylianou, D. A. & David, M. M. (2003). The nature of
scaffolding in undergraduate students' transition to mathematical proof. In the
Proceedings of the 27th Annual Meeting for the International Group for the
Psychology of Mathematics Education. (vol. 2, pp. 113-120), Honolulu, Hawaii.
7. Bransford, J.D., Brown, A.L., and Cocking, R.R. (eds) (1999). How people
learn: Brain, mind, experience and school. Washington, DC: National Academy
Press.
8. Borel, A. (March 1998). Twenty – Five Years with Nicolas Bourbaki, 1949 –
1973. Notices of the AMS, 45(3), 373 – 380.
9. Burton, L. (2004). “Confidence is Everything” – Perspectives of teachers and
students on learning mathematics. Journal of Mathematics Teacher Education, 7,
357 – 381. Netherlands: Kluwer Academic Publishers.
10. Bussiere, P., Cartwright, F., Knighton, T. (2004). The Performance of Canada’s
Youth in Mathematics, Reading, Science and Problem Solving. Canadian Results
of the OECD PISA study. Ottawa: Minister of Industry.
11. Carey, J., McCoy, D. (eds), (2006). Brain Facts: A Primer on the Brain and
Nervous System.Washington, DC: Society for Neuroscience.
12. Caroll, L. (1958). Symbolic Logic and the Game of Logic. New York: Dover
Publications inc.
13. Chapman, L.R. (ed). (1972). The process of Learning Mathematics. Pergamon
Press.
14. Cheng, P.W., Holyoak, K.J., Nisbett, R.E., Oliver, L.M. (1986). Pragmatic
versus syntactic approaches to training deductive reasoning. Cognitive
Psychology 18(3):293 – 328.
15. Clute, P.S. (1984). Mathematics Anxiety, Instructional Method, and Achievement
in a Survey in College Mathematics. Journal for Research in Mathematics
Education, Vol. 15, no. 1 (Jan. 1984), 50- 58. National Council of Teachers of
Mathematics.
16. Cobb, P., Wood, T., Yackel, E. (Jan 1992). A constructivist Alternative to the
Representational View of Mind in Mathematics Education. Journal for Research
in Mathematics Education, 23(1), 2 – 33.
 157

17. Cohen, L., Manion, L. (1980). Research   Methods in Education. Australia:

Croom Helm Ltd.
18. Connel, J. D., Langelier, C. A. (2005).   Emotions and Learning: Where brain
based research and cognitive-behavioural counseling strategies meet the road.
Rivier college online Academic Journal,   1(1), 1 – 13.
19. Cramer, D. (1997). Basic Statistics for Social Research: Step-by-step
 
calculations and computer techniques using Minitab. London: Routledge.
20. Creswell, J. W. (2005). Educational Research: Planning, Conducting and
Evaluating Quantitative and Qualitative Research second edition. New Jersey:
Pearson Education International.
21. Davis, C.S., Koch, G.G., Stokes, M.E. (1995). Categorical Data Analysis Using
the SAS System. Cary, North Carolina: SAS institute Inc.
22. Deer, G.W. (1969). The effects of teaching an explicit unit in logic on students’
ability to prove theorems in geometry. Unpublished doctoral dissertation: Florida
State University. Dissertation Abstracts International, 30, 387 – 399.
23. De Jager, T. (Dec. 1999). Do we encourage pupils to search for pattern
incorrectly? Pythagoras, 50, 20 – 24.
24. De Millo, R. A., Lipton, J., Perlis, A. J. (1979). Communications of the ACM
archive 22(5), 271 – 280.
25 Der, G., Everitt, B.S.(2002). A Handbook of Statistical Analyses using SAS
second edition. Washington, D.C.:Chapman & Hall/CRC.
26. De Villiers, M.D. (1987). Algemene beheersingvlakke van sekere wiskundige
begrippe en werkswyses deur voornemende wiskunde – onderwysers. Suid
Afrikaanse Tydskrif vir Opvoedkunde, 7(1), 34 – 41.
27. De Villiers, M. (1997). The future of Secondary School Geometry. Pythagoras,
44, 37 – 54.
28. De Villiers, M. (1990). The Role and Function of Proof in Mathematics.
Pythagoras, 24, 17 – 24.
29. De Viliers, M. (1986). The Role of Axiomatisation in Mathematics and
Mathematics Teaching. Research Unit for Mathematics Education. South Africa:
University of Stellenbosch.
30. Encheva, S.,Tumin, S.( Feb 2006).Teaching students Logical Reasoning.
Learning & Teaching Mathematics. 3, 36 – 39.
31. Edwards, L. (1997). Exploring the territory before proof: Students’
generalizations in a computer microworld for transformation geometry.
International Journal of Computers for Mathematical Learning, 2, 187 – 215.
32. Elliot, R.J. (1995). Learning SAS in the computer lab. Belmont, California:
Wadsworth Publishing Company Inc.
33. Epp, S.S. (1995). Discrete Mathematics with Applications 2nd Edition. Belmont,
Californa: Thompson Learning.
34. Epp, S.S. (Dec. 2003). The Role of Logic in Teaching Proof. American
Mathematical Monthly, 110 (10), 886 – 899.
35. Ericsson, K. A., Krampe, R. Th., & Tesch-Roemer, C. (1993). The role of
deliberate practice in the acquisition of expert performance. Psychological
Review, 100, 363 – 406.
 158

36. Garuti, R.,Boero, P., Lemut, E. (1998).  Cognitive unity of theorems and
difficulty of proof. In A. Olivier & K. Newstead (Eds.). Proceedings of the 22nd
Psychology of Mathematics Education  International Conference, 2, 345 – 352.
37. Gierdien, F. (June 2007). From “Proofs without words” to “Proofs that explain”
in secondary mathematics. Pythagoras,   65, 53 – 62.
38. Glass, G.V., Hopkins, K.D. (1978). Basic Statistics for the Behavioral Sciences.
 
Englewood Cliffs, New Jersey: Prentice Hall inc.
39. Goleman, D., (1995). Emotional Intelligence. Bantam Books: New York.
40. Gorard, S., Taylor, C. (2004). Combining Methods in Educational and Social
Research. Berkshire, England: Open University Press.
41. Griffiths, P.A. (2000). Mathematics at the turn of the millennium. American
Mathematical Monthly, 107(1), 1 – 14.
42. Hanna, G. (1991). Mathematical Proof. In Advanced Mathematical thinking
(Tall, D. ed). Netherlands: Kluwer Academic Publishers.
43. Hanna, G. (2000). Proof, Explanation and Exploration: An Overview.
Educational Studies in Mathematics, 44(1-2), 5 – 23.
44. Herbst, P. (2002). Enabling students to make connections while proving: The
work of a teacher creating a public memory. In D. Mewborn (Ed.), Proceedings of
the 23rd Annual meeting of the Psychology of Mathematics Education – North
American Chapter. Athens, G.A.: University of Georgia.
45 Herbst, P. (2002). Establishing a Custom of Proving in American School
Geometry: Evolution of the Two-Column Proof in the Early Twentieth Century.
Educational Studies in Mathematics, 49, 283 – 312.
46. Human, P.G. & Nel, J.H. With MD de Villiers, TP Dreyer & SFG Wessels.
(1989a). Alternative instructional strategies for geometry education: A theoretical
& empirical study. RUMEUS: Univ Stellenbosch.
47. Human Science Research Council. (2006). Mathematics and Science
Achievement at South African Schools in TIMMS 2003. HSRC press.
48. International Commission on Mathematical Instruction Study 19 (To be held
in Taipei in 2009). The Role of Proof and Proving in Mathematics Education.
Discussion document.
49. Johnson-Laird, P.N. (1990). The development of reasoning ability. In G.
Butterworth, & P. Bryant (eds). Causes of development (pp 85 – 110). Hillsdale,
NJ: Laurence Erlbaum Associates.
50. Jones, K. and Rodd, M. M. (2001), Geometry and Proof. Proceedings of the
British Society for Research into Learning Mathematics, 21(1), 95-100.
51. Keating, C. (2008, 23 April). Maths crisis in primary school. Cape Argus, pg 6.
52. Kleiner, I. (1991). Rigor and Proof in Mathematics: A Historical Perspective.
Mathematics Magazine, 64(5), 291 – 312.
53. Knuth, E.J. (2002). Secondary School Mathematics Teachers’ Conceptions of
Proof. Journal for Research in Mathematics Education, 33(5), 379 – 405.
54. Knuth, E.J., Elliot, R.L. (Nov. 1998). Characterizing students’ understandings of
mathematical proof. Mathematics Teacher, 91(8), 14 – 17.
 159

55.  
Kort, B., Reilly, R., Picard, R. W., (2001). An effective model of Interplay
between Emotions and Learning: reengineering educational pedagogy-building a
learning companion. In proceedings of  IEEE International Conference on
Advanced Learning Technologies. Madison
56. Kranzler, G., Moursund, J. (1995).  Statistics for the Terrified. Englewood
Cliffs, New Jersey: Prentice Hall.
 
57. Krantz, S. G., (2007). The History and Concept of Mathematical Proof.
Retrieved 14 July, 2008 from www.math.wustl.edu
58. Kutzler, B. (2000). The Algebraic Calculator as a Pedagogical Tool for Teaching
Mathematics. The International Journal of Computer Algebra in Mathematics
Education, 7(1), 5 – 23.
59. Lamport, L. (1995). How to write a proof. The American Mathematical Monthly,
102(7), 600 – 608.
60. Lauf, L. (April 2004). Reasoning and Proof in Grade 10 Mathematics. Learning
and Teaching Mathematics, 1, 23 – 25.
61. Luhanga, P. (2007, 22 May). 70% of province’s grade 3s can’t do maths, says
study. Cape Argus, p1.
62. Mason, M.M. (1998). The van Hiele Levels of Geometric Understanding. In the
Professional Handbook for Teachers: Geometry (4 – 8). Boston: McDougal-
Littell/ Houghton-Mifflin.
63. Mayfield, K.H., Chase, P.N. (2002). The Effects of Cumulative Practice on
Mathematics Problem Solving. Journal of Applied Behavior Analyis, 35(2), 105 –
123.
64. McAuliffe, S., (2007). Brain research with teaching in mind.The AMESA Western
Cape Regional Conference. Cape Town: Cape Peninsula University of
Technology.
65. McGeehan, J., (2001). Brain-Compatible learning. Green Teacher no. 364,
Spring 2001.
66. McNally, D.W.(1974).Piaget, Education and Teaching.Edinburgh: New
Educational Press.
67. Mueller, D.J. (1975). Logic and the ability to prove theorems in geometry.
Unpublished doctoral dissertation: Florida State University. Dissertation Abstracts
International, 36, 851A.
68. Page, G.W., Patton, C.V.(1991). Quick Answers to Quantitative Problems. San
Diego, California: Academic Press inc.
69. Piaget, J. (translated by Arnold Rosin). (1978). The Development of Thought:
Equilibration of Cognitive structures. Oxford: Basil Blackwell.
70. Platt, J.L. (1967). The Effect of the Use of Mathematical Logic in High School
Geometry: An Experimental Study. Doctoral dissertation. Colorado State College.
71. Reid, D. (1995). Proving to explain. In L. Meira & D. Carraher, (Eds.),
Proceedings of the Nineteenth Annual Conference of the International Group for
the Psychology of Mathematics Education, 3, 137 – 143.
72. Reid, D.(1997). Constraints and opportunities in teaching proving. In Erkki
Pehkonnen (ed.). Proceedings of the Twenty-first Annual Conference of the
International Group for the Psychology of Mathematics Education, 4, 49 – 55.
Lahti, Finland.
 160

73.   and probing: Research related to proof.

Reid, D. (2001). Proof, proofs, proving
Short Oral presentation. In M. van den Heuvel-Panhuizen (Ed.) Proceedings of
the Twentieth-Fifth Annual Conference   of the International Group for the
Psychology of Mathematics Education,(Vol. I, p. 360) Utrecht, Netherlands.
74. Reid, D. (2005) The meaning of proof  in mathematics education. Paper presented
to Working Group 4: Argumentation and Proof, at the Fourth annual conference
 
of the European Society for Research in Mathematics Education. Sant Feliu de
Guíxols, Spain. 17 - 21 February 2005.
75. Rowntree, D. (1981). Statistics without tears: An introduction for non-
mathematicians. England: Penguin books.
76. Selden, A., Selden, J. (1996). The Role of Logic in the Validation of
Mathematical Proofs. In Proceedings of the International Symposium on Teaching
Logic and Reasoning in an illogical world. Piscataway, New Jersey: Rutgers
University.
77. Skemp, R.R. (1971). The Psychology of Learning Mathematics. Middlesex,
England: Penguin books.
78. Spatz, C. (1997). Basic Statistics: Tales of Distributions 6th edition. Pacific
Grove, California: Brooks/ Cole.
79. Stillwell, J. (2002). Mathematics and its History 2nd edition. New York:
Springer-Verlag.
80. Stylianides, A. J. & Stylianides, G. J. (2008). Proof in School Mathematics:
Insights from Psychological Research into Students’ Ability for Deductive
Reasoning. Mathematical Thinking and Learning (April 2008), vol. 10, no. 2,
103 – 133.
81. Solow, D. (2005). How to read and do proofs 4th edition. John Wiley and sons.
82. Sylwester, R., (1994). How Emotions Affect Learning. Educational Leadership
52, 2:60 – 66.
83. Tall, D. (1992). Construction of Objects through definition and proof. Working
Group, Proceedings of the International Group for the Psychology of
Mathematics Education. Durham, NH: University of New Hampshire.
84. Tall, D. O.(1995).Cognitive growth in elementary and advanced mathematical
thinking. In D. Carraher and L. Miera (eds.), Proceedings of XIX International
Conference for the Psychology of Mathematics Education. Recife, Brazil. Vol. 1,
pp. 61–75.
85. Tall, D. (Dec. 1995). Cognitive development, representations and proof.
Proceedings of the conference on Justifying and Proving in School Mathematics,
27 – 38. London: Institute of Education.
86. Toulmin, S. (1958). The Uses of Argument. London: Cambridge University Press.
87. Usiskin, Z. (1982). Van Hiele levels and achievement in secondary school
geometry. (Final report of the Cognitive Development and Achievement in
Secondary school geometry). Chicago:University of Chicago, Department of
Education.
88. Volmink, J.D. (1990). The Nature and Role of Proof in Mathematics Education.
Pythagoras, 23, 7 – 10.
 161

89. Vygotsky, L., (Cole, M., John-Steiner,   V., Scribner, S., Souberman, E.,
(eds).(1979). Mind in Society: The Development of Higher Psychological
 
Processes-. Massachusetts: Harvard University Press.
90. Walter, R.L. (1972). The effect of knowledge of logic in proving mathematical
theorems in the context of mathematical  induction. Unpublished doctoral
dissertation: Florida State University. Dissertation Abstracts International, 33,
 
262A.
91. Weber, K. (2003). Students’ Difficulties with Proof. MAA Research sampler 8.
Retrieved 19 July 2007, from http://www.maa.org/t.
92. Yevdokimov, O. (2003). The Place and Significance of the Problems for Proof in
Learning Mathematics. In Proceedings of the 3rd Conference of the European
Society for Research in Mathematics Education. Italy: Bellaria.
93. Young, L. (2003). Retrieved 19 July 2007 from
www.math.ksu.edu/math791/finalpaper/laneeterm.
 162

 
APPENDICES
 
Appendix A1
 

UNIVERSITY OF THE
  WESTERN CAPE
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

Read through the following puzzle and then try and solve it. Please do the puzzle on your
own and we will allow time at a later stage to discuss your solution with your fellow
students. Write your solution in the space provided. Explain in your own words the
reasoning that you used to arrive at your solution.

PUZZLE I

In the back of an old cupboard you discover a note signed by a pirate famous for his
bizarre sense of humour and love of logical puzzles. In the note he wrote that he had
hidden treasure somewhere on the property. He listed five true statements (a – e below)
and challenged the reader to use them to figure out the location of the treasure.

a) If this house is next to a lake, then the treasure is not in the kitchen.
b) If the tree in the front yard is an elm, then the treasure is in the kitchen.
c) This house is next to a lake.
d) The tree in the front yard is an elm or the treasure is buried under the flagpole.
e) If the tree in the backyard is an oak, then the treasure is in the garage.

Where is the treasure hidden?

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 163

 
Appendix A2
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

Read through the following puzzle and then try and solve it. Please do the puzzle on your
own and we will allow time at a later stage to discuss your solution with your fellow
students. Write your solution in the space provided. Explain in your own words the
reasoning that you used to arrive at your solution.

PUZZLE III (POST TEST)

Your grandfather that is known for his sense of humour and love of logical puzzles left
you a note. In the note he wrote that he had hidden your birthday present somewhere on
one of his properties. He listed five true statements (a – e below) and challenged you to
figure out the location of the present.

a) If this house is next to a main road, then the present is not in the attic.
b) If there is a swing in the yard then the present is in the study.
c) The yard has a lawn or the present is in the cupboard next to the stove.
d) If the yard has a lawn, then the present is in the attic.
e) The house is next to a main road.

Where is the present hidden?

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 164

 
Appendix B1
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

Read through the following puzzle and then try and solve it. Please do the puzzle on your
own and we will allow time at a later stage to discuss your solution with your fellow
students. Write your solution in the space provided. Explain in your own words the
reasoning that you used to arrive at your solution.

KNIGHTS AND KNAVES (pre-test)

The logician Raymond Smullyan describes an island containing two types of people:
knights who always tell the truth and knaves who always lie. You visit the island and are
approached by two natives who speak to you as follows:

A says: B is a knight
B says: A and I are of opposite type.

What are A and B?

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 165

 
Appendix B2
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

Read through the following puzzle and then try and solve it. Please do the puzzle on your
own and we will allow time at a later stage to discuss your solution with your fellow
students. Write your solution in the space provided. Explain in your own words the
reasoning that you used to arrive at your solution.

KNIGHTS AND KNAVES (post - test)

The logician Raymond Smullyan describes an island containing two types of people:
knights who always tell the truth and knaves who always lie. You visit the island and are
approached by two natives C and D but only C speaks.

C says: Both of us are knaves.

What are C and D?

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 166

 
Appendix C1

 
Arguments with quantified statements (pre-test)
 
Rewrite the following statements in formal language
 
29.
1. No birds, except ostriches, are nine feet high.
2. There are no birds in this aviary that belong to anyone but me.
3. No ostrich lives on mince pies
4. I have no birds less than nine feet high.

30. 1. All writers who understand human nature are clever.

2. No one is a true poet unless he can stir the hearts of men.
3. Shakespeare wrote Hamlet.
4. No writer who does not understand human nature can stir the hearts of
men.
5. None but a true poet could have written Hamlet.
 167

 
Appendix C2

UNIVERSITY OF THE  WESTERN CAPE

DEPARTMENT OF MATHEMATICS AND APPLIED MATHEMATICS
 
APPLIED MATHEMATICS 112
CLASSTEST 6
 
DATE: 5 APRIL 2007

Arguments with quantified statements (post-test)

Rewrite the following statements in formal language

1 No bank closes before 3:30 unless it is a small bank.

2. No shark eats plankton unless it is a whale shark.
3. Students never study unless they have to prepare for a test.
4. None but a true gentleman will offer his seat to a lady on a bus.
5. None but a brave soldier will fight in a war
 168

 
Appendix D1
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

Reorder the premises in the following argument to make it clear that the conclusion
follows logically. It may be helpful to rewrite some of the statements in if – then form
and to replace some statements by their contrapositives.

ARGUMENTS WITH QUANTIFIED STATEMENTS

PUZZLE I (pre – test)

11. When I work a logic example without grumbling, you may be sure it is one I
understand.
12. The arguments in these examples are not arranged in regular order like the ones I
am used to.
13. No easy examples make my head ache.
14. I can’t understand examples if the arguments are not arranged in regular order like
the ones I am used to.
15. I never grumble at an example unless it gives me a headache.

∴ These examples are not easy.

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 169

 
Appendix D2
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

Reorder the premises in the following argument to make it clear that the conclusion
follows logically. It may be helpful to rewrite some of the statements in if – then form
and to replace some statements by their contrapositives.

ARGUMENTS WITH QUANTIFIED STATEMENTS

PUZZLE II (post – test)

16. There is no box of mine here that I dare open.

17. My writing-desk is made of rose-wood.
18. All my boxes are painted, except what are here.
19. There is no box of mine that I dare not open, unless it is full of live scorpions.
20. All my rose-wood boxes are unpainted.
∴ My writing-desk is full of live scorpions.

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 170

 
Appendix E1
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

SET THEORY
(pre-test)

For all sets A, B and C prove the following:

A ∩ ( B ∪ C ) = ( A ∩ B) ∪ ( A ∩ C )

……………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 171

 
Appendix E2
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

SET THEORY
(post-test)

For all sets A, B and C prove the following:

( A ∪ B) ∪ C = A ∪ ( B ∪ C )

……………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 172

 
Appendix F1
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

METHOD OF DIRECT PROOF AND DIVISIBILITY

(pre-test)

Prove the following by using the method of direct proof:

For all integers a, b, and c, if a / b and a / c then a / (b+ c)

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 173

 
Appendix F2
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

METHOD OF DIRECT PROOF AND DIVISIBILITY

(post-test)

Prove the following by using the method of direct proof:

For all integers a, b, and c, if a / b and a / c then a / (b – c)

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 174

 
Appendix G1
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

METHOD OF DIRECT PROOF I

(pre-test)
Prove the following statements using the method of direct proof:

1
(i) For all integers m, if m 〉 1, then 0〈 〈1
m
(ii) The difference of any two rational numbers is a rational number.

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 175

 
Appendix G2
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

METHOD OF DIRECT PROOF I

(post-test)
Prove the following statements using the method of direct proof:

(i) For all real numbers x, if 0 〈 x 〈 1, then x2 〈 x

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 176

 
Appendix H1
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

METHOD OF INDUCTION
(Pre-test)

Use mathematical induction to prove that:

2 + 4 + 6 + …2n = n2 + n,
for all integers n ≥ 1
………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 177

 
Appendix H2
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

METHOD OF INDUCTION
(Post-test)

Use mathematical induction to prove that:

1 + 5 + 9 + …(4n – 3) = n(2n – 1),
for all integers n ≥ 1
………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 178

 
Appendix I1
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

STRONG MATHEMATICAL INDUCTION

(RECURSIVE SEQUENCES)
(PRE-TEST)

1 Suppose that c0 , c1 , c 2 , ... is a sequence defined as follows:

c0 = 2, c1 = 4, c 2 = 6,
c k = 5 c k −3 for all int egers k ≥ 3
Prove that c n is even for all integers n ≥ 0

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 179

 
Appendix I2
 
UNIVERSITY OF THE WESTERN CAPE
 
DEPARTMENT OF MATHEMATICS AND APPLIED
MATHEMATICS
 

COURSE: APPLIED MATHEMATICS 112

NAME & SURNAME:…………………………………………………………….

STUDENT NO:……………………………. DATE:………………………

FIRST LANGUAGE:…………………………….

STRONG MATHEMATICAL INDUCTION

(RECURSIVE SEQUENCES)
(Post-test)

1 Suppose a1 , a 2 , a 3 , ... is a sequence defined as follows:

a1 = 1, a 2 = 3,
a k = a k − 2 + 2a k −1 for all int egers k ≥ 3.
Prove that a n is odd for all integers n ≥ 1

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………

………………………………………………………………………………………………
 180

 
Appendix J

 
QUESTIONAIRRE FOR NCS TEACHERS
2007
 

 
Degree/ Diploma and state where it was obtained:…………………………………….

Highest qualification in Mathematics:…………………………………………………

School :…………………………………………………………………………………

1. What teaching approach (methodology) do you use when teaching mathematical

proofs?

…………………………………………………………………………………….
2. Are you aware of other teaching approaches besides the one you are using?
(yes or no). If you are aware of other approaches, name them.

…………………………………………………………………………………….
3. What type of proof is used in the grade 10 to 12 mathematics curriculum?

…………………………………………………………………………………….
4. Do you know of other types of proof besides the proof that is used in the school
curriculum? If you know other types, name them.

…………………………………………………………………………………….

…………………………………………………………………………………….
5. Have you studied mathematical logic or do you know about mathematical logic?
Please specify.

…………………………………………………………………………………….

…………………………………………………………………………………….

6. If you did study mathematical logic or if you know about it do you know where in
school proofs logic is used? If you do know, name at least one example.

………………………………………………………………………………………
 181

 
Appendix K

 
PARTICIPANT CONSENT FORM
 
VERIFICATION OF ADULT INFORMED CONSENT FOR OWN
PARTICIPATION
 

I,…………………………………………………………………………………………….
(Please print full name and surname)

voluntarily give my consent to serve as a participant in the study entitled:

PROVING MATHEMATICAL STATEMENTS: BY FIRST YEAR STUDENTS

AT THE UNIVERSITY OF THE WESTERN CAPE.

I have received a satisfactory explanation of the general purpose and process of this
study, as well as a description of what I will be asked to do and the conditions that I will
be exposed to.

It is my understanding that my participation in this study is voluntary and I will receive

no remuneration for my participation.

It is further my understanding that I may terminate my participation in this study at any

time and that any data obtained will be held confidential. I am aware that the researcher
has to report to his supervisor and that all data collected will be accessible to the
supervisor as well.

Signature of participant:…………………………………………………………………….

Date:………………………………………………………………………………………..
 182

 
Appendix L