TOPIC THREE

FUNDAMENTALS OF MEASUREMENT AND EVALUATION

INTRODUCTION
The unit focuses on what measurement (evaluation) and its uses in the context of education.
Measurement (evaluation) is brought out by analyzing keys terms namely, data (numeric
information), variable, continuous and discrete variable, descriptive and inferential
statistics.
In the unit the scales (levels) of measurement are also discussed, which should be of concern
to any teacher (researcher) for it dictates the method of statistics (i.e. statistical analytical
method), which can be used with data. This subject of determining the correct statistical
method to use for particular data is not dealt with here for it is beyond the scope of the unit.

LEARNING OUTCOMES
At the end of the unit the learner should be able to:
1. Define variable, continuous and discrete variable, data, statistics, measurement,
descriptive and inferential statistics.
2. Explain why educators need to have rudimentary (basic) knowledge of statistics.
3. Identify the relationship between statistics and measurement.
4. List the four levels of measurement.
5. Describe the characteristics (properties) of the 4 levels of measurement.
6. Associate the levels of measurement with any data collected by a teacher (researcher).

Rationale for doing statistics and measurement

Most of the journal articles in social sciences (psychology, communication, education,
sociology etc.) report findings in statistical form or apply statistical concepts in their
discussions or research reports. Consequently, we need at least rudimentary (basic)
knowledge of statistics in order to be intelligent consumers of such research. Furthermore, if
we intend to do research in our area of interest, we must understand statistics in order to use
and plan appropriate procedures and to interpret and communicate the findings from our
research. So statistics is needed for these purposes i.e. in:
i) Research- extension of knowledge and solving problems.
ii) In interpreting mass of data (test scores or marks, sales, etc.)
 Considerations possible with data or the two ways of treating data:

Data

Descriptive statistics:
Data treated as if they are coming from a
population. No inferences are made. This Inferential statistics:
deals with a mass of numerical data to be Conclusions about population
interpreted by organizing and summarizing are made using a representative
(compiling) these data in a way that they finite sample. Thus, in this we
can be understood and communicated make reasonable decisions with
without generalizing beyond the group incomplete information i.e.
under consideration. taking a small finite group
(usually a random sample) and
then inferring or generalizing.
Thus one uses a sample
(random) to study the
population in inferential
statistics.

Measurement has to be involved in obtaining data.

Measurement and types of variables

Measurement refers to the assigning of numbers to
individuals or objects in a systematic way as a means
of representing properties of the individuals (or
objects).
The properties, features or characteristics of the
individuals or objects that are quantified during
measurement are usually referred to as a ‘variable’
which is possessed by different individuals or objects
in different quantities.
 A variable is a characteristic that can be
measured and that can assume different
values. Height, age, income, province or country
of birth, grades obtained at school and type of
housing are all examples of variables. Variables
may be classified into two main categories:
categorical (qualitative variables) and
numeric (quantitative variables).

Quantitative and Qualitative variables

In statistics, a variable is an attribute of the sample population that is being studied. It
can vary from member to member of the sample, and it can be just about anything.
Color, size, height, weight, and any other attribute that can be measured or described
can be a variable. There are two general types of
variables: qualitative and quantitative. Qualitative variables can be further broken
down into nominal, ordinal, and binary, while quantitative variables can be broken
down into discrete and continuous.

The Qualitative variable definition in statistics is one that falls into descriptive
categories, things that are not measured. Color, taste, smell, style, mood, and
satisfaction are examples of qualitative variables. These variables are not measured
with numbers but rather describe the quality of something. For example, height (6 feet
4 inches) is not a qualitative variable but size (he is tall) is. Qualitative variables are
also called categorical variables. As mentioned, there are several different
types.Nominal Variable
i. Nominal variable
Uses descriptive categories, but there is no inherent order to them. For instance,
naming of class streams such as Form One A, Form One B, Naming of Dormitories as
Kilimanjaro, Longonot, Mt.Kenya or types of vehicles, Mercedes Benz, Toyota,
Volkswagen, Range Rover etc fall into categories, but there is no order to those
categories. Which one should be first? Other examples of nominal variables are

i. Degree classifications.
ii. School categories as National, Extra County, County and Sub-county.
iii. Gender.-male female, boys and girls.
iv. School categories as Boarding, Day, Mixed, Single gender etc.
v. School categories on location-Urban and Rural Schools

i. Ordinal Variable
An ordinal variable has an intrinsic order. They have a rank, size, placement, or
other sorts of order that matter. Categories of pizza size, for instance, can be ordered
from smallest (personal pan) to largest (NY style mega). Some other examples of the
ordinal variable are

a) Military ranks such as Major General, Colonel, service men etc.

b) Calendar days Monday, Tuesday, Wednesday, Thursday, Friday, Saturday,
Sunday.
c) Year in school, form one , form two, form three, form four or year one, year two
year three, year four.
d) School sizes, large, small, medium.
e) Business sizes, large, small, micro-business.

One important thing to remember is it is not necessarily an even amount between two
categories of an ordinal variable. In a corporate structure, for instance, the jump up
one level (in terms of salary or responsibility) from the receptionist to salesperson
might not be the same as the jump up one level from sales to management.

Binary Variable
A binary variable, as the name suggests, has only two options. Examples of binary
variables are

i. Yes or no
ii. True or false
iii. Mother or father
iv. Positive or negative

Examples of Qualitative Variables

Qualitative variables are in categories that are descriptive, not measured. In other
words, they vary by quality.

a) Flavor
b) Hat shape
c) Hairstyle
d) Vegan or non-vegan
e) Crispy or grilled chicken
f) Favorite movie
g) Olympic gymnastics events

The quantitative variable definition in statistics is one that can be measured and
assigned a numerical value. It describes a quantity instead of a quality.
Quantitative variables are also called numerical variables. Height, weight, age,
speed, diameter, and the number of marbles in a bag are all examples of quantitative
variables.
Types of Quantitative variables

These include;

i. Discrete Variable.
A discrete has a set number of values that are allowed over a range. For instance,
when counting the number of people who watched movie A versus movie B, the
variable will have a numerical value, but only whole numbers. There will never be
47.51 people in a theater. Over the range of possible values, this variable is limited to
a discrete set. Provide measurements in whole numbers. For example 10 people, 34
cows, 10 Presidents. There are no fractions in discrete variables. Other examples of
discrete variables are

i. Number of cows.
ii. Population of people in a county or country.
iii. Number of students in class.

ii. Continuous Variable

A continuous variable is not limited like a discrete variable. Though it may be
restricted to a range, a continuous variable includes every value in that range. Height
is a continuous variable because a person can be 150 cm tall, 150.5 cm, 150.555, and
so on, the limits of the measuring device notwithstanding. Other examples of
continuous variables are

a. Weight-89.5Kg
b. Volume of water in a glass 13.2 Litres
c. Time to run a marathon 2.45minutes
d. Distance to various towns, Nyeri to Nairobi 150KM.
e. Number of bags of maize= 125.5bags.
f. Temperature=27.9 degrees Celsius.

NB: Data expressed in decimals and fractions.

 s

Figure 1: Types of variables.

What is Level of Measurement?

In statistics, level of measurement is a classification that relates

the values that are assigned to variables with each other. In
other words, level of measurement is used to describe
information within the values. Psychologist Stanley Smith
Stevens (1946) is known for developing four levels of
measurement: nominal, ordinal, interval, and ratio.

Four Measurement Levels

The four measurement levels, in order, from the lowest level of

information to the highest level of information are as follows;

1. Nominal scales

Nominal scales contain the least amount of information. In nominal

scales, the numbers assigned to each variable or observation are
only used to classify the variable or observation. For example, a
fund manager may choose to assign the number 1 to small-cap
stocks, the number 2 to corporate bonds, the number 3
to derivatives, and so on.
These categorize variables in terms of their kind and
numbers used have no meaning. Numbers used are only for
identification.

i. Gender- Boys and girls, men and women.

ii. Categories of clasess-Form One, Form two, Form Three, Form

Four

iii. School categories-National, Extra County, County, Sub-County.

iv. Names of dormitories-Kilimanjaro, Longonot, Dr Krepf.

v. Names of Counties- Nyeri, Mombasa, Garissa etc./ County

number 1, county number 2, county number 3…..

vi. Types of Universities- Public and Private, Arts based and Science
based etc.

2. Ordinal scales

Ordinal scales present more information than nominal scales and

are, therefore, a higher level of measurement. In ordinal scales,
there is an ordered relationship between the variable’s
observations. For example, a list of 500 managers of mutual
funds may be ranked by assigning the number 1 to the best-
performing manager, the number 2 to the second best-performing
manager, and so on.

With this type of measurement, one can conclude that the number
1-ranked mutual fund manager performed better than the number
2-ranked mutual fund manager.

In education, ordinal scale provides an order that can be

used for comparison purposes. Numbers used cannot be
subjected to mathematical computations but bring out the
magnitude for comparison purposes.

1. Greater than or less than: Form one A ha greater mean score

than Form one B.

2. Degree classifications-First class, second class, etc.

3. Rankings- Highest to lowest academic cadres- professor,

associate, senior lecturer, lecturer, tutorial fellow etc.

3. Interval scales
Interval scales present more information than ordinal scales in that
they provide assurance that the differences between values are
equal. In other words, interval scales are ordinal scales but with
equivalent scale values from low to high intervals.

For example, temperature measurement is an example of an

interval scale: 60°C is colder than 65°C, and the temperature
difference is the same as the difference between 50°C and 55°C. In
other words, the difference of 5°C in both intervals shares the same
interpretation and meaning.

Consider why the ordinal scale example is not an interval scale:

A fund manager ranked 1 probably did not outperform the fund
manager ranked 2 by the exact same amount that a fund manager
ranked 6 outperformed a fund manager ranked 7. Ordinal scales
provide a relative ranking, but there is no assurance that the
differences between the scale values are the same.

A drawback in interval scales is that they do not have a true zero

point. Zero does not represent an absence of something in an
interval scale. Consider that the temperature -0°C does not
represent the absence of temperature. For this reason, interval-
scale-based ratios fail to provide some insights – for example, 50°C
is not twice as hot as 25°C.

Zero does not mean absence of the variable. For example in

education a student who scores 0% does not mean has zero
knowledge but the meaning is that he or she was not able to
answer any item correct in the exams given.

4. Ratio scales

Ratio scales are the most informative scales. Ratio scales provide
rankings, assure equal differences between scale values, and have a
true zero point. In essence, a ratio scale can be thought of as
nominal, ordinal, and interval scales combined as one.

For example, the measurement of money is an example of a ratio

scale. An individual with $0 has an absence of money. With a true
zero point, it would be correct to say that someone with $100 has
twice as much money as someone with $50.

Application: In education a student scores 60% in Geography while

another scores 95% in Geography, the difference in the two
students is thus 95%-60%=35%