Professional Documents
Culture Documents
Pyc2616 10050590
Pyc2616 10050590
PYC2616/1/2024-2030
10050590
InDesign
PR_Tour_Style
CONTENTS
Page
...........
iii PYC 2616/1
...........
iv
INTRODUCTION
A word of welcome
A very warm welcome to you as you begin this module called Psychological
survey development PYC2616.
which the questionnaire to be designed will be used. This implies that when
designing a questionnaire for research or assessment purposes, the population
characteristics should be considered in order to produce a questionnaire that
is scientifically sound, fair and unbiased, and does not infringe on the rights
of the users. This is important, especially when working in a multicultural
context such as South Africa with a rich history of discriminatory practices
of psychology emanating from the apartheid and colonial rule. Therefore,
designing a questionnaire that is psychometrically sound means that you have
to be mindful of the influence of Western approaches that inform questionnaire
development as this will help you to identify acceptable and unacceptable
practices in the development of a questionnaire. Additionally, this will also deter
you from using a one-size-fits-all approach in designing a questionnaire. Thus,
enabling you to design a questionnaire that is reliable and valid, conceptualised
from a multicultural perspective, and with the consciousness of the context in
which it will be utilised.
...........
vii PYC 2616/1
INTRODUCTION
...........
viii
I nt r o du c ti o n
...........
ix PYC 2616/1
INTRODUCTION
Course structure
The following foundational competence is embedded within the module,
and will be assessed directly or indirectly through assessment of the specific
outcomes in terms of the following criteria:
Study aids
To help you study, the following are provided:
There are activities in each lesson. These include questions to think about,
self-tests, exercises to do, and drawings to interpret. They are designed
to help you to draw out important ideas and main points. Occasionally,
they are designed to test how carefully you have worked through the
material. We strongly urge you to complete these activities, since they
are an important aid to helping you understand the texts you will be
studying. Model answers to questions and self-tests can be found at the
end of each lesson.
...........
xii
I nt r o du c ti o n
It is important to know where you are going before you start studying. By
studying this module, we want to get to a certain point and help you to focus
on this goal. We must plan the route and make sure that we take the correct
paths to lead you there. In academic terms, we call the destination (end goal/
place where we want to go to) the outcomes of the module. The road we travel
to get there is called assessment criteria. Like information boards next to the
highway, these criteria help us to make sure that we are on the right path, and
they also show us how far we have to travel. If the criteria state that there are
eight check points on this road and we can make ticks against seven of those
(because we know the contents, can apply them to our own situation and are
able to lead a debate about this subject), we know that we are nearly reaching
our goal – just one more issue to learn, interpret and implement in our context
and we will have completed the task.
Feedback
You will receive detailed feedback on Assignments 01 and 02. You will therefore
be able to evaluate your learning progress and prepare for the examination by
taking heed of the feedback written by the lecturer. Answers to the questions
as well as discussions of the questions are provided. The key to each question
is indicated, that is, the option that provides the correct answer. The reason
why the indicated key is the most correct answer is explained. In some cases,
the discussion also indicates why the other options are incorrect. Check your
answers and see where you make mistakes. Then go back to the sections in
your study guide.
Thematic approach
As explained, this study guide contains the required resource material and
lessons. You must work through the lessons and complete the necessary
actions. Note that the activities follow on each other. If you complete all the
activities, you can expect to obtain the desired skills and knowledge. You
will then be able to apply the competencies you acquired in developing your
own questionnaire and writing a manual for your questionnaire to evaluate a
...........
xiv
I nt r o du c ti o n
A final word
Now that you have received the above information, we hope that you feel
more confident to start with the lessons.
References
Claasen, N. C. W. (1997). Cultural differences, politics and test bias in South
Africa. European Review of Applied Psychology, 47, 297–307.
Clay, A. (2017). Decolonizing psychology in South Africa. American
Psychological Association, 48(10). https://www.apa.org/monitor/2017/11/
decolonizing-psychology
Dudgeon, P., & Walker, R. (2015). Decolonising Australian psychology:
Discourses, strategies, and practice. Journal of Social and Political
Psychology, 3(1), 276–297.
Furo, A. (2018). Decolonizing the classroom curriculum: Indigenous knowledges,
colonizing logics, and ethical spaces [Doctoral dissertation, University
of Ottawa]. University of Ottawa Research. https://ruor.uottawa.ca/
handle/10393/37106
Laher, S., & Cockcroft, K. (2014). Psychological assessment in post-apartheid
South Africa: the way forward. South African Journal of Psychology,
44(3), 303–314.
Van de Vijver, FJR & Leung, K. (1997). Methods and data analysis for cross-
cultural research. Newbury Park, CA: Sage.
Van de Vijver, FJR & Poortinga, YH. (1997). Towards an integrated analysis
of bias in cross-cultural assessment. European Journal of Psychological
Assessment, 13, 29–37. doi:10.1027/1015-5759.13.1.29.
Van de Vijver, FJR & Tanzer, NK. (2004). Bias and equivalence in cross-cultural
assessment: An overview. Revue Européenne de Psychologie Appliquée/
European Review of Applied Psychology, 54, 119–135. doi:10.1016/j.
erap.2003.12.004.
...........
xv PYC 2616/1
INTRODUCTION
...........
xvi
1 LESSON 1
This lesson deals with the identification of a content domain for a questionnaire.
Aims
In lesson 1, you will learn the definition of a content domain in the context of
psychology and social science research and experiments. You will be given
activities with examples to further your knowledge.
Lesson
• content domain
• suitability of a questionnaire
• shared norms: socially desirable behaviours (e.g. “the do’s and do nots”)
• shared beliefs: ideas or assumptions about the world
• shared values and expectations: moral standards perceived as desirable and
esteemed
...........
1 PYC 2616/1
tionnaire will in turn give you ideas about the content domain, that is, all the
possible tasks, behaviours, attitudes, etcetera implied by the purpose. It is from
this content domain that you select items to make up the questionnaire.
1.1 INTRODUCTION
Let us first talk about characteristics of a good questionnaire and survey.
A good test has a clearly defined purpose. To define the purpose, the test
developer and the test user must be familiar with these questions: what is the
test supposed to measure; what is the domain or the content it is measuring;
who will take the test; who is the test intended for; and how will the test score
be used. Is the test designed to compare the performance of test takers to each
other or are they designed to diagnose. Different types of items and scores are
used for these different types of comparisons.
So, let us begin with an overview of experimental validity and content domain.
This is just one example, but every experiment, test or study has a content
domain that it is trying to accurately measure.
1.2 GLOSSARY
Here are some basic terms related to the topic of identifying the content domain
for a questionnaire that you may not be familiar with, but which you need to
know. They will come up in your reading of the material in this lesson.
...........
3 PYC 2616/1
...........
4
1 AC TIVITIES
Before you work through the activities, I want you to read the resource
material for this lesson (content domain and suitability of a questionnaire).
Suitability of a
Content domain questionaire
Press the button Press the button
or go to page 112 in or go to page 141 in
the study guide. the study guide.
T ime re cord
These activities took me ........ hours.
interest in a topic. In this case, you can read the first three actions but you do
not need to do them.
AC T I O N 01
Make two lists, one list of the things that interest you and another of the things
that you do. These lists can be as long as you like but they should contain at
least five items in each category.
1. 1.
2. 2.
3. 3.
4. 4.
5. 5.
AC T I O N 02
Rank the items on your interest list from the most interesting to the least
interesting and rank the items on your activities list (things you do) from the
things you spend most time on to the things you spend least time on.
AC T I O N 03
Inspect your rank-ordered lists. You can probably identify common themes
(that is, things that the lists have in common) from these lists. These should
give you a clearer indication of your world and the things that are important
to you. Based on your interests and activities, now make a list of things you
would like to find out or get more information on that would be of use to you
or someone else. Complete this sentence:
Now you are ready to identify a topic and describe the content domain for
your questionnaire.
AC T I O N 0 4
You have already identified a general theme of interest, either through completing
the first three actions or because you already knew at the outset what you
want to do. Even if your ideas are vague, write your ideas in a sentence that
begins as follows:
...........
6
AC T I O N 05
Because your topic is fairly broad, you now need to narrow it down. Select a
problem area within the topic that you want to investigate and write it down.
AC T I O N 0 6
Once you have narrowed down the topic, you need to reduce it even further
to give it some focus for your investigation. Look at what you have written and
decide on some specific questions about the problem area within the topic
you have identified. Write them down.
Carefully consider the questions you have written. Do they all relate to each
other? Can you combine them into one question? Should you rather choose
one question and leave the others for separate questionnaires? When you have
made these decisions, you have the focus of your questionnaire. Note that these
questions help you to be specific about what you want to measure with your
questionnaire; they are not the final items or questions for the questionnaire,
which will only be formulated in lesson 3.
AC T I O N 07
Now that you have identified the focus for your questionnaire, you are ready
to limit the scope of the questionnaire so that you cover the content adequately
but do not ask irrelevant questions. Look at your questions (the previous action)
and write what you need to know in order to get the relevant information that
will provide the answers to your questions. For example, if one of your questions
was, “Do younger people perceive crime differently from older people?”, you
would need information on the respondents’ ages (biographical information)
and perceptions of crime.
Now ask yourself whether each point you have written is really relevant to the
focus of your investigation and give a reason/reasons for including each one.
If the answer is “yes” and you can provide a reason for including each point,
you have successfully limited the scope of your questionnaire. If the answer is
“no”, cross out the ones that are not relevant. If there are some points that sound
interesting but you cannot provide reasons for including them, cross them out
too. This exercise should have helped you to successfully limit the scope of
your questionnaire. (Did you realise that the information you have identified
as being relevant is actually the content domain of your questionnaire? You
Activity 1.3 will select your items according to the content domain.)
Study the resource
material
Now that you have limited the scope of the topic and described your content
Resource domain, you have to make sure that a questionnaire is the right type of measuring
suitability of a instrument to use for obtaining the kind of information you want. Study the
questionnaire
resource on the suitability of a questionnaire as a measuring instrument.
...........
7 PYC 2616/1
AC T I O N 01
Activity 1.4 You need to describe the reason for choosing a questionnaire as your measuring
Evaluate the instrument. In order to decide whether a questionnaire is the most suitable
suitability of a measuring instrument for your investigation, it is important to understand why
questionnaire as
a measure for this and when questionnaires are used. Complete the following sentence:
content domain
Questionnaires are used for …
AC T I O N 02
Now you can match your content domain with a questionnaire as the measuring
instrument of choice. If you have not already done so, write the purpose of
your questionnaire and describe the content domain. To do this, you have to
formulate or describe what it is that you want to investigate and that includes
the topic you are interested in. For example, the purpose of your questionnaire
might be to investigate young people’s perceptions of how crime influences
their future plans. The content domain would be any information relevant to
this topic (of perceptions of how crime influences future plans) and the tasks,
behaviours and attitudes, etcetera, for obtaining this information. Now that
you have identified the content domain for your questionnaire, complete the
following sentence:
Now go through the exercise described in your resource material again, that
is, decide whether you can use a questionnaire to obtain the information you
need (or whether you would have to administer ability tests and so on). If the
answer is yes, a questionnaire will be a suitable measuring instrument for your
investigation. If not, see if you can focus on a different aspect of your topic
that is suited to a questionnaire.
Once you have identified the content domain for your questionnaire, you can
go on to the next step of deciding what your questionnaire should look like.
In lesson 2, you will look at the questionnaire specification document.
S E L F -A SS E SS M E N T
Study the scenario carefully before you consider the self-assessment questions.
Lifestyle refers to the experiences and events in a person’s life. A style of living
is a pattern of experiences and events. It includes the principles according to
which a person lives, and it deals with the different areas of the person’s life
such as his/her home and work environments. A health and fitness centre
decided to include the option of a stress management programme as part
of the service given to their clients. Suppose you were contacted to evaluate
...........
8
QUESTION 1
Pule: What is the content domain represented by the lifestyle questionnaire?
You agree most with, second most with and least with:
QUESTION 2
The main reason why a questionnaire is the most suitable measure for this
content domain is because it ...
(1) A, C, E and G
(2) B and C
(3) B, D and G
(4) All of the above.
...........
9 PYC 2616/1
QUESTION 3
Operationalising lifestyle in terms of specific areas (home environment, work
environment, nutrition and fitness, and recreational activities) helps to define
the ... of the questionnaire.
Q U E S T I O N 1: O P T I O N 1 I S T H E K E Y – P U M U LO, ZO L A A N D
TEMBA
The content domain reflects the purpose of a questionnaire. This questionnaire
covers behaviours in different areas of a person’s life, such as personal
relationships, the work environment and physical health. These aspects all
relate to the larger content domain, which is factors influencing lifestyle. This
is why Pumulo’s answer is the best answer. If the content domain were the
assessment of lifestyle, then all the items would be concerned with various ways
or methods for assessing lifestyle. Since this is clearly not the case, Temba is
incorrect. Zola’s response is also not correct – while there are some items that
relate to stress, the focus of the questionnaire is not stress management either.
Q U E S T I O N 2: O P T I O N 2 I S T H E K E Y – B A N D C
The choice of measuring instrument depends on the purpose of the investigation
and the content to be measured. You will see in your resource material (see:
suitability of a questionnaire as a measuring instrument) that the main purpose
of a questionnaire is to gather factual information. In this instance, the aim is to
gather information about lifestyle (B) but not about stress (A). A questionnaire
also provides a way to get information for further analysis (C). This is why option
2 provides the correct answer. A questionnaire does not provide a means for
investigating a relationship between variables (D), nor does it measure specific
abilities (F). This particular questionnaire covers some aspects of lifestyle, not
all of them (G), and it does not deal with stress management (E).
Q U E S T I O N 3: O P T I O N 4 I S T H E K E Y – S CO P E
In this instance, the topic (option 3) or general theme of interest is lifestyle. The
content domain (option 1) of the questionnaire relates to some of the factors
that affect lifestyle. By identifying these factors, you have defined the scope
of your questionnaire. This is why option 4 provides the correct answer. The
actual nature of these factors or aspects will determine what kind of questions
you use (option 2).
...........
10
Lesson
2 2
2 Design a questionnaire
This lesson helps you to create a simple survey questionnaire based on principles
of effective item types and layout.
AIMS
Lesson
• content domain
• item format
• layout of a questionnaire
• specification document
...........
11 PYC 2616/1
can create avenues around communication barriers that can lead to frustration
that can develop between the survey research, researchers, and the participants.
Whether or not we are aware of it, “culture” and its influences are ever constant
in all that we do. In striving to find the richness of culture, both our own and of
the population we serve, we as researchers are able to discover ourselves, our
participants and the context of their life circumstances. In this module, there is
an increased awareness to do away with the notion that “culture” is exotic and
intrinsic to only those who are from groups other than our own. As each of us
let go of the mindset that culture is an unimportant trend, we sharpen in on an
essential and often overlooked key ingredient of effective psychological survey
development. We are able to “know” that consideration of culture and its influ-
ences is an integral ingredient to effective psychological survey development
and no survey can achieve success without a well-designed questionnaire that is
sensitive to culture and diversity, and prevent ethnic discrimination. In addition,
the relevant norms may vary across social classes or subcultures within a society.
When there is such variation in norms, the bias induced by socially desirable
responding may distort the observed associations between the behaviour in
question and the characteristics of the respondents, besides affecting estimates
of overall means or proportions. Therefore, do not allow cultural differences to
become the basis for criticism and judgments and avoid stereotypes. Mahatma
Ghandi said: “no culture could live if it attempts to be exclusive.”
2.1 INTRODUCTION
The heart of any survey research project is the survey questionnaire itself. A good
questionnaire helps to achieve the research objectives; it provides complete and
accurate information, is easy for the respondents to complete, is so designed
as to make sound analysis and interpretation possible and is brief.
The distinct steps in designing a good questionnaire are: decide on the information
required; define the target respondents; select the method(s) of reaching the
...........
12
...........
13 PYC 2616/1
...........
14
...........
15 PYC 2616/1
...........
16
2.2 GLOSSARY
Here are some basic terms related to the topic of designing a questionnaire that
you may not be familiar with, but which you need to know. They will come
up in your reading of the material in this lesson.
...........
17 PYC 2616/1
Item scaling Scales are used for Skale word gebruik vir Dilekanyetsi di
Itemskaal rating performance, prestasiegradering, sebedisetswa ho
indicating aanduiding lekanya mosebetsi,
Ho lekanya dintho preferences, and van voorkeure ho bontsha dikgetho
expressing opinions en menings en le ho hlahisa
and attitudes. houdings. maikutlo.
Layout of a The questionnaire Die vraelysuitleg Tlhophiso ya
questionnaire layout or format of -formaat sluit lenanepotso e
Uitleg van ’n vraelys includes organising die uitleg van die kenyeletsa ho
the cover page, voorblad, instruksies hlophisa leqephe la
Tlhophiso ya instructions vir die voltooiing sekoahelo, ditaelo tsa
lenanepotso for completing the van die vraelys, die ho tlatsa lenanepotso,
questionnaire, the vraagvolgorde en die tatelano ya dipotso le
question sequence algehele vraelysuitleg tlhophiso kaofela ya
and the overall in. lenanepotso.
questionnaire layout.
Questionnaire A list of the ’n Lys van die vereiste Lenane la ditlhokwa
specification the required eienskappe vir die tsa lenanepotso.
document characteristics for vraelys. Eintlik moet Ha e le hantle,
Vraelysspesifi- the questionnaire. enige inligting lesedi lefe kapa lefe
kasiedokument Basically, any rakende die werklike le potapotileng
information vraelys en waarom lenanepotso la nnete
Tokomane ya ditlhoko surrounding the die verskillende vrae le hore hobaneng
tsa lenanepotso actual questionnaire ingesluit is (of nie) dipotso tse fapaneng
and why the different ingesluit word in die di kenyeleditswe
questions were spesifikasiedokument. (kapa ha di a
included (or not) kenyeletswa) di
should be included lokela ho kenngwa
in the specification tokomaneng ya
document. ditlhoko.
2 AC TIVITIES
Before you work through the activities, I want you to read the resource
material for this lesson (content domain, item format, layout of a
questionnaire and specification document).
Layout of a Specification
questionaire document
Press the button Press the button
or go to page 127 in or go to page 140 in
the study guide. the study guide.
...........
18
T ime re cord
These activities took me ........ hours.
AC T I O N 01
Identify different types of items and scaling methods. When you read the
resource on the layout of the questionnaire, you will see that it is important
to have a balance of different types of questions in order to maintain the
respondents’ interest and desire to complete the questionnaire as well as to
collect all the relevant information. In the resource on item format, different
types of items (or questions) are described. Look at the following items and
in the right hand column, write down what kind of item each one is. Indicate
whether it is an open question, closed question (and of which specific type)
or a rating scale (and again of which type).
Item Type
1. Do you have a valid driver’s licence? Yes/no
2. Why do people need to have a valid driver’s licence?
3. People should have a driver’s licence (choose one answer)
for identification purposes
to prove that they can drive
in case they have an accident
...........
19 PYC 2616/1
ANSWER
The item types are:
(1) Closed question, because there is a limited choice of answers (yes or no).
(2) Open question, because it requires the respondents to state their own
opinions and allows for any kind of answer.
(3) Closed question, because there is a limited choice of answers (multiple-
choice type).
(4) Closed question, because there is a limited choice of answers (true/false
type).
(5) Rating scale, semantic differential type (extreme scale points are opposing
adjectives).
(6) Closed question, because there is a limited choice of answers (checklist/
inventory).
(7) Rating scale, Likert type.
(8) Open question, because it allows the respondents to give any kind of
answer.
(9) This looks like an open question but it is actually a closed question because
it implies only a yes or no answer. It would be an open question if you
asked, “What is your opinion about increasing the driving age to 21 years?”
...........
20
AC T I O N 02
Link the item format and scaling method to the purpose and content of your
questionnaire. In lesson 1 you decided on the purpose and content domain
of your questionnaire. Now you need to decide what kind of items to use in
order to get the information you want. The uses, as well as advantages and
disadvantages of different types of items, are discussed in the resource on
item format.
To get an idea of how to link item format and scaling method to the purpose
and content of the questionnaire, let us use the example of the questionnaire
to investigate whether South Africans experience crime-related stress, referred
to in the resource on identifying the content domain for a questionnaire. In the
blank spaces below, write down what sort of questions (that is, item format)
you would use to get the information you need. The first two are filled in as
an example.
To find out the respondents’ age, you might just ask an open question like “How
old are you?” but some respondents might find this offensive or insensitive,
particularly older respondents. If your questionnaire is for adults, it might be
better to offer them a choice of answers, such as age groups. For example, you
could say, “Indicate your age group by marking the applicable box below”:
The question of gender is easier because the options are obviously limited. A
closed question will suffice.
...........
21 PYC 2616/1
Check through your list to make sure that you have included a balance of
different types of items. Make sure that you have some closed questions, some
open questions and at least one rating scale.
Activity 2.3 The length of a questionnaire is not random. It has to be considered carefully
Decide on the total to ensure that you get the information you want but do not lose respondents
number of items because it is too long or boring. This topic is discussed in the resource called,
Resource layout of the questionnaire. The first step is to identify the extent to which each
Layout of a
content area (the information you need) needs to be covered; the second step
questionnaire is to consider the characteristics of your respondents and the time available
for testing.
...........
22
AC T I O N 01
You need to identify the coverage required for each content area. One of the
considerations in deciding on the number of items for your questionnaire is
the number of different smaller content areas or constructs covered by your
questionnaire. In the previous activity, you used the example of an investigation
of South African’s experiences of crime-related stress and for this investigation,
five content areas were identified. These were biographical information,
personal experience of crime, levels of stress, reactions to different crimes, and
the effect of crime on daily lives. You therefore need at least one item on each
of these content areas. In some cases, one item is not enough. For example,
if you want information on stress levels associated with different crimes, you
might want to use a rating scale. Rating scales do not have a fixed number of
items (or things to be rated) but for the purposes of this assignment, your rating
scale should consist of at least twelve items. It is also useful to have more than
one item dealing with the same aspect to serve as a control so that you can
see whether the respondent is answering questions consistently or not (see
the resources on “reliability” and “validity”). For example, in addition to your
rating scale, you might also have an open-ended question that deals with the
same content area.
In the previous activity, you made a list of the information you need to investigate
the topic. In this way, you have in fact already identified your smaller content
areas that make up the focus of your investigation. In the space below, write
down your list of smaller content areas and next to each one, make a note of
the coverage required. The most important particular content area may require
more than one item whereas less important areas will only require one item.
The first one has been completed as an example. You can also use the example
of the questionnaire on crime as a guideline.
You have now identified the coverage required but you may have to modify it
because of the type of respondents who will complete the questionnaire and
the time they typically will have available.
AC T I O N 02
You should evaluate the impact of characteristics of respondents and the time
available for completing the questionnaire. Read the guidelines in length of
the questionnaire in your resource on the layout of the questionnaire. Now
write down the most important factors to take into account when deciding on
the length of a questionnaire.
...........
23 PYC 2616/1
You could probably cover the content domain comprehensively with 21 items
(some of which may be grouped into a rating scale containing approximately 12
items). We could break down the coverage of the content areas as follows: the
first three items would be closed questions to collect biographical information
(first content area), then a filter question (closed, yes/no type) followed by an
open question on personal experience of crime (second content area), a rating
scale (consisting of 12 items) on levels of stress associated with different crimes
(third content area), a closed (multiple-choice) question on personal reactions
to crimes and an open question to serve as a control (fourth content area), an
open question on perceptions of the effect of crime (fifth content area) and
lastly an open question for any other comments the respondent may wish to
add. We therefore have five closed items, four open items and a 12-item rating
scale (total of 21 items).
If you want to involve people from all strata of society as respondents for
your questionnaire, remember to make your items straightforward. Your
respondents may not be specialists but rather the person in the street, therefore
the questionnaire should not be too long or complicated. Most people lead
busy lives and do not want to spend much time completing the questionnaire.
The average person should be able to answer these items in half an hour.
Some of the items may be answered very quickly (for example, the questions
on biographical information or the checklist), while others may take a little
longer (open-ended questions, for instance).
Now you are ready to apply this exercise to your own questionnaire. Using
the previous exercise as a guideline for your own questionnaire and looking
at the different types of items as well as the coverage of content areas you
decided on earlier, write how many items you think you would need for the
entire questionnaire. Also, list the content areas and the number of items for
each one. The first item is filled in as an example.
...........
24
Now that you have decided on the length of your questionnaire, you need to
consider the layout or the way in which your items will be presented.
Activity 2.4 First impressions are important! This means that the layout of your questionnaire
Design the layout is important for conveying the idea that it is worthwhile completing the
of the questionnaire questionnaire. Nobody wants to bother with a sloppy document that is not
Resource professional. In activity 2.4, you will be given the opportunity to work on the
Layout of a layout of your questionnaire.
questionnaire
AC T I O N 01
Provide an introduction and instructions for completing the questionnaire. By
referring to the resource on the layout of the questionnaire, write a suitable
introduction to your questionnaire.
Introduction: ……………………………………........................…………………………..
Of course you can include anything you want to in your introduction but it is
suggested that you include at least the four main points provided in the layout
of the questionnaire resource.
AC T I O N 02
Now organise your items in a logical and appropriate sequence. When you
read through the resource on the layout of the questionnaire, you will see that
the presentation and sequence of items are important. Before you decide on
your own questionnaire, let us practice on the example we worked on in the
previous section, that is, the investigation of South Africans’ experience of
crime-related stress. You have already identified approximately 21 items (in
the previous exercise) for this questionnaire. Now organise these items in a
logical sequence, using the guidelines in your resource material (presentation
and sequence of questions in the layout of the questionnaire resource). Start
by asking yourself whether there are any sensitive or leading questions and
whether you need a filter question. Once you have identified these, it makes
the organisation easier.
Our suggestion would be to start with a broad question that orients the
respondent to the topic, such as an open-ended question on people’s perceptions
of the effect of crime in South Africa. This could be followed by the 12-item
...........
25 PYC 2616/1
rating scale for stress levels associated with different types of crime. (Can you
see that we are moving from the general to the more specific?)
We might then introduce the topic of personal experience of crime and use a
filter question followed by an open question about people’s personal experience
of crime. Personal reactions may be regarded as a sensitive topic and therefore
these items should be placed towards the end of the questionnaire. You might
want to use a closed question and an open question to serve as a sort of validity
check for this content area.
Now that you have worked through the example, you are ready to tackle your
own questionnaire. Using the guidelines provided here, look carefully at the
items you identified for your own questionnaire and then provide the sequence
in which you think they should be presented.
S E L F -A SS E SS M E N T
Study the scenario carefully before you consider the self-assessment questions.
...........
26
QUESTION 1
Ben: Do you think the instructions that the AIQ provides to respondents
are correct and complete?
Temba: The instructions provide the correct information but are incomplete.
Bets: I think the instructions provide the correct information and they
are complete.
You agree most with, second most with and least with:
QUESTION 2
Question 3.1 in the scenario is an example of which type of item format? It
is a/n ...
...........
27 PYC 2616/1
QUESTION 3
You want to ask respondents whether they have used recreational drugs but
you realise that this may be a sensitive question. Where is the best place to
put a sensitive question like this?
Q U E S T I O N 1: O P T I O N 5 I S T H E K E Y – FAU N A , T E M B A A N D
BETS
The instructions are incomplete because they do not refer to the intention of
the questionnaire and the matter of confidentiality. There is also a problem
with the information that is provided. It asks the respondents to respond to
each question by ticking an answer. Some questions cannot be ticked.
Please note that questionnaire instructions refer to how the questionnaire should
be completed. It does not refer to the instructions provided with each question.
Q U E S T I O N 2: O P T I O N 2 I S T H E K E Y
Item 3.1 is an open question as respondents are required to provide their own
“personal” answers. Closed questions (such as item 2.1) provide respondents
with a limited choice of alternate answers. In item 3.1 the respondents do not
receive a choice of alternative answers, and therefore this item is not a closed
question.
Q U E S T I O N 3: O P T I O N 4 I S T H E K E Y
Sensitive or very personal issues should be near the end of the questionnaire.
The respondents, if asked to respond to such questions in the beginning, may
be unwilling to complete the rest of the questionnaire.
2.3 REFERENCES
Prabhu, RK. & Rao, UK. (1960). Mind of Mahatma Gandhi. India: Navajivan Trust
...........
28
Lesson
3 3
In this lesson, you will learn more about the process of writing questionnaire
items.
AIMS
Lesson
...........
29 PYC 2616/1
One of the ways to help resolve this is to asked the people concerned how they
would like to be referred to, thus making them the centre of your item devel-
opment. Your questionnaire should consist of items that are not foreign to the
people you are studying and the researcher should not impose a particular way
of understanding to his/her participants.
3.1 INTRODUCTION
In this lesson, you will be taken through the process of writing questionnaire
items. You will learn about the important things you need to take into
consideration when writing the items and what you need to avoid.
a colonial lens where only western ways of knowing are prioritised. Most
research studies still target black communities and western assumptions are
imposed in the way data is gathered.
These are the aspects we are urging you, as a student, to be critical of. It is
always important to ask yourself whose reality am I using to measure this
experience. The problem with the two questions is that western concepts
are used to measure experiences of people in non-western contexts and the
result is the production of knowledge which pretends to be of indigenous
people. What decolonial researchers such as Chilisa (2012) propose is to allow
indigenous people to express their experiences through their own methods.
when using the conceptualisations of the western world when studying non-
western communities. In doing so, they distorted those communities’ cultures
and presented them as backward and resistant to change. As widely reported,
no culture can be perceived as static but changing one’s ways of doing should
not be imposed. The western education has expected other nations to assimilate
its cultures instead of evolving on their own.
3.2.2 Language
Another critical aspect, which is connected to the issue of culture mentioned
above, is the need to consider language when developing questionnaire items.
Researchers are always advised to use simple and clear language. It is important
to remember that often times your respondents will come from different walks
of life and may not be familiar with the terms that social science researchers
think are basic. Researchers must therefore avoid using jargon (language only
understood by those in one’s field of work).
• Leading questions – these are the types of questions that can sway the
respondent in a particular direction. They are also called suggestive questions.
The respondent might get a hint on how you want them to answer the
question. For example, “Isn’t it safer being at school than at home?”
• Double-barrelled questions – these questions require more than one answer.
The problem with it is that the respondent would not know which part
of the question to respond to and the researcher would have difficulty in
analysing responses from such questions. For example: “Does the university
consider race and gender in its admission policies?” An answer could apply
to only one variable and not the combination.
• Avoid using presumptions – this applies in instances where the researcher
assigns participants into particular categories. An example of a presumptuous
question would be: “What type of cereal do you eat in the morning?” The
researcher has already assumed that the respondents eat cereal in the
morning.
• Ambiguous questions – these are the types of questions that could have
multiple meanings. The respondents could interpret the questions in many
different ways. For example, if you ask students “Are you satisfied with
the curriculum?” the students could be satisfied with some aspects of the
curriculum and dissatisfied with others. The researcher will also not reap
the intended rewards from the question as the responses will not help him/
her to work on the aspects of the curriculum that require an improvement.
...........
32
would have the same questionnaire in three versions. This is to ensure that
everyone is responding in the language he or she truly understands. However,
this does not resolve the problem of the culture of the academy we introduced
in the last section. The variables being measured are still formulated with a
western lens. If the research is measuring the level of education, education will
still be based solely on western terms. It will not include indigenous definitions
and formulations. It therefore becomes important to rather use language to get
an in-depth understanding of the knowledge these languages carry.
• What can we learn from indigenous languages about indigenous ways of
being?
• What do these languages tell us about their systems?
These are the questions future studies need to help us understand if we are
to truly deal with the remnants of the colonial past embedded in the ways of
doing research.
...........
33 PYC 2616/1
3.4 GLOSSARY
Here are some basic terms related to aspects of questionnaire item writing
discussed above. They will aid your understanding of the material on this topic.
...........
34
3 AC TIVITIES
Before you work through the activities, I want you to read the resource
material for this lesson (writing questionnaire items).
Writing
questionnaire items
Press the button
or go to page 147 in
the study guide.
AC T I O N 01
METHOD In Lesson 02 you specified the content areas for your questionnaire together
Activity 3.1
with the number of items and type of items for each content area. You have
also indicated the sequence of presentation of these items. Keeping in mind
Write items for a
questionnaire the criteria discussed in the previous activity, you now need to write the items
for your questionnaire as specified in Lesson 02.
Resource
Writing
questionnaire
items
...........
35 PYC 2616/1
AC T I O N 02
Evaluate these items. Do they all comply with the general principles for
constructing new items? If they do, then well done. If some, or all, of your items
breach one or more of the general principles, then, (a) write each ‘incorrect’
item, in its original form (b) indicate why the item is ‘bad’, and (c) reformulate
the item so that it accords with the general principles for constructing ‘good’
items.
AC T I O N 03
You can now compile a draft version of your questionnaire consisting of an
introduction (refer to Lesson 02) and the items you have written (and corrected
where necessary). Note that with the rating scale you need to indicate the
rating associated with each number on the questionnaire. Examples are: 1 =
disagree entirely up to 5 = agree entirely; 1 = strongly disagree up to 5 =
strongly agree; 1 = never up to 5 = always.
S E L F -A SS E SS M E N T
QUESTION 1
Below is a list of questionnaire items. Please look at them carefully and based
on what you have learnt above indicate what is problematic about the way
there are formulated.
A. Most children start their schooling at the age of six years. Were you also
six years old when you started school?
B. How far are you from the nearest hospital and how often do you visit it?
C. Are you satisfied with township life?
D. What is your favourite fizzy drink?
...........
36
QUESTION 2
Do you remember that we cautioned you to be mindful of the culture the
questionnaire items are communicating? Can you explain how a particular
culture is communicated and presented as a norm in the first two questions?
QUESTION 1
• The first question is a leading question. The respondent is led into a particular
direction.
• The second question is a double-barrelled question. It is asking two things
at once.
• The third question is ambiguous, it is too broad, and it does not ask about
specific aspects of township life.
• The fourth question is presumptuous; the researcher has assumed that the
respondent takes fizzy drinks.
QUESTION 2
• As we indicated before, formal schooling is a western method of education.
This excludes people who acquire their learning outside the schooling
system. The first question is therefore communicating a western viewpoint.
• The second question is measuring only one form of a healthcare facility,
which is a hospital. It is also communicating a western cultural point of view.
It is only asking about one form of healthcare, which is a hospital. There
could be different types of healthcare facilities that are not of western origin.
...........
37 PYC 2616/1
Lesson
4 4
This lesson deals with pilot-testing the initial version of the questionnaire.
AIMS
In this pilot-testing phase of developing your questionnaire, you will learn how
to distinguish between strong and weak items. You will be guided by activities
and examples through which you will administer the questionnaire and you will
make use of item analysis procedures to revise the questionnaire.
Lesson
This lesson will help you to produce a set of items to be included in the final
version of the questionnaire.
• correlation coefficient
• item analysis
• writing questionnaire items
that the concepts and words you are using will not be imposed on others or that
the concepts you are using are not foreign to other people’s language, culture
and lived experience and conditions. When the questionnaire is constructed,
you must make sure that you do not portray other people’s ways of living and
their experiences as superior to others (i.e. placing others’ way of being and
concepts at the centre and translating from those experiences for other people).
The sample used for piloting will be able to assist you in identifying biased con-
tent or items from their responses. This would help in assessing whether your
questionnaire is addressing what you want to measure. This will be a step towards
refining content and concepts in your questionnaire in order to produce good
quality information and to provide you with the best possible understanding
of the topic you are investigating.
4.1 INTRODUCTION
At this stage of the questionnaire development, you will learn how to assess
your items, by assessing the validity of your concept/content of the items of
the questionnaire in relation to its sample and with the use of certain statistical
tools (i.e. item analysis). Through this process, you will be able to distinguish
between good/strong items and weaker items. By weeding out weak items,
you get a better quality measurement of the issue you intend to investigate –
similar to focusing the lenses of a pair of binoculars. The pilot-testing of the
questionnaire also gives you an assurance that the information you gather
is actually dealing with the topic you want to investigate. You want to gain
assurance that the data you gather when you eventually administer the final
...........
39 PYC 2616/1
4.2 GLOSSARY
Below are brief explanations of some basic terms you will encounter in this
lesson. Acquaint yourself with their meaning in order to have a better grasp
of the learning content.
...........
40
4 AC TIVITIES
It is important that you read the resource material for this lesson (correlation,
item analysis and writing questionnaire items).
Correlation Writing
coefficient Item analysis questionnaire items
Press the button Press the button Press the button
or go to page 114 in or go to page 118 in or go to page 147 in
the study guide. the study guide. the study guide.
T ime re cord
These activities took me ........ hours.
AC T I O N 01
You need to identify a suitable sample. You already have a draft questionnaire
and an introductory talk, but to actually administer your questionnaire (i.e., to
try it out), you must get hold of people to try it out on! Like any other sampling
situation, the sample should represent the population to which you hope to
generalise your findings.
You have already described the target population for your questionnaire (the
population that your questionnaire is intended for), but you may want to have
...........
41 PYC 2616/1
another look at what you did in lesson 1. Do you still agree with how you
described your target population then, or do you want to change anything?
Write a short description of how you now see your target population.
In an ideal world, you would be able to select your sample from the target
population using sophisticated sampling techniques such as stratified random
sampling. However, for the purposes of pilot-testing a questionnaire, researchers
rarely go to such great lengths to find a truly representative sample. Instead,
they simply get together a sample that appears to be as similar as possible to the
target population. So if your questionnaire is meant for both men and women
and for people with a university degree, try not to choose a sample consisting
only of men who have never been to university. If your questionnaire is meant
for managers, do not use a sample consisting of manual workers.
You do not need hundreds of people for pilot-testing your questionnaire. For
purposes of item analysis it is usually said that one needs at least one more
person than there are items in one’s scale. Since you want to analyse your
rating scale and there are 12 items in your rating scale, that means you need
at least 13 people. We can relax this rule a bit and say you need between 10
and 15 people in your sample.
Now give some thought as to where you will get hold of such a sample. You
could ask your friends and family to try out the questionnaire, provided they
are similar to your target population. Another good starting place is where
you work or study – maybe you could ask suitable people there to do the
questionnaire. Your church or any other organisation to which you belong
can also be a good place to find a sample. You need not administer your
questionnaire to everybody in the sample at the same time, and not everybody
in the sample need to come from the same place.
Having given the practicalities of getting hold of a sample some thought, write
some notes about the place or places where you expect to get your sample from.
...........
42
AC T I O N 02
Administer the questionnaire to the sample. If you are lucky, you will be able
to get all the people in your sample together in one place (for example, they
may be a group of school children) so that they can all do the questionnaire at
the same time. However, you may have to get people to do the questionnaire
one at a time or in small groups. In whatever way you decide to administer
the questionnaire, be sure to be ethical about what you are doing. Use the
following checklist to make certain that you are acting ethically. The answer
to each of the questions should be YES.
When you administer the questionnaire, it is best to find a quiet place where
people will not be disturbed and to allow them as much time as they need. It is
usually a good idea if you read the introduction at the top of the questionnaire
out loud – especially if you have a group of people together. Then answer
any questions they have. While they are completing the questionnaire, glance
at what each person is doing to ensure that they have the right idea of what
is required, but do not look over people’s shoulders the whole time – some
people find it very irritating! Be available to answer any questions people may
have. Of course you cannot tell people the right answer to any question since
there is no right answer and it is up to them how they want to respond, but
you can help them where something seems unclear. Keep notes of the kind
of questions people ask and what difficulties arise – you can use these notes
later to make improvements to your questionnaire.
Summarise the notes you took while administering the questionnaire. These can
include, for example, places where the instructions were unclear, notes about
particular items that seem to be giving people trouble, and notes about any
unusual events that occurred while you were administering the questionnaire.
When you get the questionnaire back, it is a good idea to quickly scan it to see
that they have completed all of it. People often inadvertently leave out some
...........
43 PYC 2616/1
parts (for example, biographical information), and the less missing information
you have to deal with the better. If anything is missing, ask the person if he
or she would mind quickly completing that part of the questionnaire as well.
AC T I O N 03
There are two ways of using a pilot study to improve your questionnaire. The
first is to use what happened during the study (for example, people asking
questions about some of the items, or people getting confused about the
instructions) to revise the questionnaire. Look again at the notes you made
while you were administering the questionnaire and then see where you can
improve the questionnaire. Were the instructions clear enough? Were there
some questions that people did not quite understand? Were there questions
that offended some people? Now write a short summary of what changes (if
any) you have decided to make to your questionnaire.
Activity 4.2 The second way of using a pilot study to improve a questionnaire is to analyse
Do an item analysis the responses people gave to each item in the questionnaire. This is known as
item analysis. Some professional test constructors use sophisticated statistical
Resource
procedures in their item analyses, but even without any statistical tools, it is
Item analysis
possible to do the two most essential types of item analysis: remove items with
Writing the too little variance and remove items that do not discriminate (see the resource
questionnaire
items on item analysis). To keep things simple, you only need to perform an item
correlation analysis on the rating scale part of your questionnaire, but it is possible to do
coefficient similar analyses on all kinds of items.
AC T I O N 01
Compile a data sheet. You should focus on the 12 items in the rating scale part
of the questionnaire. If you constructed the rating scale as suggested in previous
lessons, the possible responses to each item in the scale will have a number from
1 to 5 and people will have ticked a box corresponding to one of the numbers
for each item. For example, here is how a particular respondent completed
the first three items of a rating scale measuring “noise tolerance”.
Statement 1 2 3 4 5
Never Almost Some- Most Always
never times of the
time
I like loud music. X
I prefer quiet places. X
I enjoy noisy environments. X
The ticked options are known as item responses. This respondent’s item
responses suggest that he or she is somebody who is very tolerant of noise as
he/she always likes loud music, enjoys noisy environments most of the time
and never likes quiet places.
...........
44
Statement 1 2 3 4 5
Never Almost Some- Most Always
never times of the
time
I like loud music. X 5
I prefer quiet places
X 5
(Reverse score)
I enjoy noisy environments. X 4
It is difficult to work with item responses and scores spread out over 10 or 15
questionnaires (one for each respondent), so you need to collect all of them
together in one place. If you know how to use a computer, you could type
the scores into a spreadsheet program or a statistics package. However, if you
are not computer literate, you can use a data sheet like the following. At this
stage your data sheet should have 12 items and 15 cases. Unfortunately, we
cannot show a full 12 items and 15 cases here because the space is limited.
This example has five items and eight cases. Simply expand this table to
accommodate your 12 items by 15 cases scenario.
Items Total
Cases
1 2 3 4 5 score
1
2
3
4
5
6
7
8
...........
45 PYC 2616/1
The data sheet is divided into rows and columns – one row for each person
in your sample and one column for each item in your rating scale. There is
space on the data sheet for eight people and five items, but as your sample is
larger you have to add rows and columns to accommodate the additional items
and cases. You will notice that there is an extra column for a case number
(from 1 to 8) and total score. Case numbers are helpful for finding particular
questionnaires in case you make a mistake when writing down data, while
total score is the score a person got on the scale as a whole.
Now take a questionnaire from your pile of questionnaires. Write the number
“1” on the front of the questionnaire – this is the case number. Next, go to
the rating scale part of your questionnaire and transfer the score for each
item to the first row on the data sheet. (Remember to reverse-score where
necessary before transferring the scores.) When you have finished with the
first questionnaire the first row on the data sheet will be completely filled in
(except for the block under “total score”). Take the next questionnaire from
your pile, write the number “2” on it and fill in the second row on your data
sheet. Continue like this until you have transferred all the item scores from
the rating scales to the data sheet. Below is an example where the first three
item scores of the first case (person) who completed a “noise tolerance” scale
have been filled in. The person got item scores of 5 for item 1, 5 for item 2
and 4 for item 3.
Items Total
Cases
1 2 3 4 5 score
1 5 5 4
2
3
4
It is easy to make mistakes while transferring the item responses from the
questionnaires to the data sheets, so when you have finished doing it for
everybody in your sample, check that you have not made any mistakes. Take
two or three questionnaires from your pile and check carefully that you wrote
the item responses correctly on the data sheet. If you do not find any errors,
good! If you do, go back and check every questionnaire.
Finally, calculate each respondent’s total score on the scale and write it in the
“total score” column on the far right of the data sheet. You get the total score
by adding up the person’s item scores. Below is an example where the total
scores for the first two people in a sample have been calculated. The lowest
possible total score anybody can have on the rating scale is five (if they got
an item score of 1 for each of the five items), while the highest possible total
score is 25 (if they got an item score of 5 for each of the five items).
Items Total
Cases
1 2 3 4 5 score
1 5 5 4 3 4 21
2 3 2 2 1 3 11
3
4
...........
46
Choose two or three of your cases at random and check if you have calculated
their total scores correctly. If you find any errors, redo all the calculations.
AC T I O N 02
You are now ready to start with the actual item analysis of your rating scale
(see the resource on item analysis). The first step is to find items with too little
variance. “To vary” means to change; and items with too little variance are
those items where almost everybody in the sample gets the same item score
(i.e. almost everybody answers the item in the same way). (See the resource
on item analysis.) You want your scale to show differences between people,
for example if you have a noise tolerance scale, you want to know who has a
high noise tolerance and who does not. Items where everybody gets the same
score do not help you to see differences between people and are therefore of
no use in a scale. On the other hand, items where there is much variation (or
variance) in how people respond help you to see differences between people.
Of course most items are somewhere in between, for example one might find
the following situation with an item: nobody gets an item score of 5, a few
people get item scores of 1, 2 or 4 and most people in the sample get an item
score of 3. Would you say such an item shows enough variance?
Items Total
Cases
1 2 3 4 5 score
1 5 1
2 3 3
3 3 2
4 3 4
5 4 1
6 3 3
7 3 4
8 3 2
By running your eye down the column of item scores under item 1 and then
the column of item scores under item 2, you should be able to see that item
2 has more variance and is therefore the better item. Most people responded
in the same way to item 1 (their item response was 3), whereas there is much
more variation in response to item 2. Item 2 may not be perfect, for example,
...........
47 PYC 2616/1
nobody had an item response of 5, but it is certainly a better item (in terms
of variance) than item 1.
There are statistical methods for showing which items in a scale show the
least variance and should therefore be discarded, but you are not expected
to calculate any statistics. Simply do the same for all the items in your scale
as you did for the two example items above – that is, run your eye down
each of the columns on your data sheet and look for items that may not have
sufficient variance. If a column contains mainly only one number (e.g. almost
all 4’s with very few other numbers), the item does not show much variance.
If a column contains a good spread of numbers (for example many 2’s, 3’s and
4’s and some 1’s and 5’s), the item shows lots of variance. If a column contains
mainly one number, but with a fair number of other numbers as well, the item
shows some variance.
In this way, you identify what looks like the worst items in your scale in that
they seem to have less variance than the other items. Now write down the
numbers of the three or four worst items in your questionnaire. Then, after
examining the way each item was worded, write why you think it may be that
the item has so little variance.
You now have a list of items that may be problematic due to a lack of variance.
Next, you will add items to the list that are problematic for a very different
reason. Then you will decide which ones to get rid of.
AC T I O N 03
To measure differences between people our items need to show some variance,
but even if the items show lots of variance, the scale may not measure anything
in particular. What if each item measures something completely different
(one item measures noise tolerance, another item measures attitude towards
mathematics, etc) so that it becomes impossible to know what the scale as a
whole measures?
You can do something to ensure that each item in the scale measures more
or less the same thing and that the items are not too divergent. This will
...........
48
guarantee that you can sensibly add each of the item responses together to get
a person’s score on the scale as a whole. How will you know if each of the
items broadly measures the same thing? One method is to check if people’s
responses on each item correspond (or “correlate”) with their responses to
the scale as a whole. The idea is that people who get a high score on an item
should also get a high score on the scale as a whole, while those who get a
low score on the item should get a low score on the scale as a whole. If an
item can “discriminate” in this way between high and low scorers on the scale
as a whole, it is a good item. One normally talks of discrimination as a bad
thing, but in this case, you want an item to discriminate between high and low
scorers because it shows that the item measures more or less the same thing
as the other items in the scale.
Suppose the following were the scores a sample of five people got for item 3
in an “optimism” scale, together with the total scores they got for the scale as
a whole. (Note that only five cases and five items are shown here.)
Items Total
Cases
1 2 3 4 5 score
1 4 53
2 1 12
3 5 49
4 4 40
5 2 16
Would you say that item 3 discriminates well between high and low scorers
on the scale as a whole? Why?
Item 3 does seem to be pretty good at discriminating between high and low
scorers. Respondents who get high item scores (4’s and 5’s) also get high total
scores (40 and over), while those who get low item scores (1 or 2) also get low
total scores (under 20). Looking whether item scores correspond with a total
score, is called item-total correlation and it is a technique often used in item
analysis. Professional questionnaire constructors usually calculate a correlation
coefficient (see the resource on correlation coefficient). A correlation coefficient
is an index of how strongly two variables are related. The idea is to establish
how strong each item-total correlation is. But here you are not expected to
do this. Instead, you can use a simple way of getting an impression of how
strongly each item correlates with the total.
Simply make a small scatterplot for each item showing the relationship between
each item and the scale total. Scatterplots are explained in the resource on the
correlation coefficient. Here is a scatterplot showing the relationship between
item 3 and the scale total for the optimism scale discussed above.
...........
49 PYC 2616/1
Each dot on the scatterplot represents a person. The dot is positioned (roughly)
next to the correct total scale score on the left and above the correct item score
at the bottom. As you can see, there were five people in the sample (five dots)
and the dots are arranged roughly in a diagonal line from bottom left to top
right. This means that there is a strong correlation between the item score and
the total score – that is the item discriminates well. It is unlikely that you will
find correlations nearly this strong, but if the dots for your sample are arranged
in a vaguely similar pattern, then you will know that there is at least some
correlation (the item discriminates). If the dots do not seem to form a pattern
at all, then there is no correlation (the item does not discriminate), and if the
line seems to go from top left to bottom right, there is a negative correlation
(the item does discriminate, but the wrong way round, so it is no good).
If your questionnaire has 12 items, you will have to draw 12 scatterplots (one for
each of the 12 items in your scale), and if you administered your questionnaire
to 15 individuals you will have 15 dots on each scatterplot (one for each person
in your sample). You will also have 12 total scores (one for each person in your
sample). Do not worry about being too precise, so long as your scatterplots
show roughly how well each item correlates with the scale total.
As was the case for item variance, there are also no absolute rules with regard
to item discrimination, but we advise you not to be too strict. In the social
sciences, correlations are often not quite as strong as one would like them
to be, and if you insist that each item should correlate very strongly with the
scale as a whole, you may end up having no items left at all! Considering all
your scatterplots (12, if you have 12 items), identify what appears to be the
worst items in your scale in terms of failure to discriminate. Now write down
the numbers of the three or four worst items. Then, after looking again at the
way each item was worded in your questionnaire, write why you think it may
be that the item discriminates so badly.
The reason why items do not discriminate is usually because they measure
something different from the other items in the scale. Sometimes one can see
why this is the case when one looks again at the wording of the item, but
sometimes it seems inexplicable and one just has to accept that it is so.
AC T I O N 0 4
...........
50
You are now ready to compile a final version of your questionnaire. You have
to make some tough choices regarding which items to keep and which to
discard from your scale. Your scale should have eight items, so if you started
with 12 items you can afford to discard four items. Study the two lists of bad
items you compiled – one list for items that do not show much variance and
the other list for items that do not discriminate well. You may find that some
items appear on both lists. Decide which four items should be discarded and
write down their numbers, together with a short justification as to why you
chose to get rid of each of these items.
There is no fixed rule about identifying your worst items. If you find that you
really cannot decide between two items (maybe one that shows little variance,
but does discriminate somewhat versus one that shows a lot of variance but
hardly discriminates at all), simply flip a coin. If you do not seem to have
enough really bad items to discard, find some that are a little less good than
the others. Getting rid of the worst items will not give you a perfect scale,
but it will almost certainly be a better scale than if you had not performed an
item analysis and just kept all the items. Most importantly, your final scale is
more coherent (the items all measure broadly similar things) than the original
scale – which means (as you will see in the next lesson) that it has a greater
degree of reliability.
Finally, taking into account the notes you made during the pilot-testing of the
questionnaire, revise the instructions to the questionnaire, change whatever
items you think need changing and remove the worst items from the rating
scale part of the questionnaire. You have now pilot-tested your questionnaire
and have used the pilot-test to construct a final version of the questionnaire.
S E L F -A SS E SS M E N T
QUESTION 1
Consider the responses below of four participants to a questionnaire consisting
of five items:
...........
51 PYC 2616/1
(1) A and C
(2) B and D
(3) A, B and C
(4) A,B, C and D
(5) None of the above.
QUESTION 2
Based on the table above, select the correct statement(s):
(1) A
(2) B
(3) A and B
(4) None of the above.
QUESTION 3
Which of the statements below do you agree with:
(1) A and B
(2) B and C
(3) A and C
(4) A, B and C
(5) None of the above.
...........
52
QUESTION 4
Please study the graph below and answer the following questions.
(1) A and B
(2) A, B and C
(3) A, C and D
(4) B and C
(5) B, C and D
Q U E S T I O N 1: O P T I O N 3 I S T H E K E Y – A , B A N D C
Items 2 and 3 have comparable spreads of numbers, which is larger than any
of the other items. They are therefore also better discriminators between high
and low scores than any of the other items. Item 5 shows the smallest variance
and is therefore the most ineffective item of the group.
Q U E S T I O N 2: O P T I O N 3 I S T H E K E Y – A A N D B
Item 3 is a good discriminator since it shows a good correspondence between
a high item score and a high score for the scale as a whole. In addition, low
item scores correspond well with low total scores.
Item 5 is the weakest item since it shows the least correspondence between
the item and total scores.
...........
53 PYC 2616/1
Q U E S T I O N 3: O P T I O N 3 I S T H E K E Y – A A N D C
Statement A is correct. The item score should be reverse-scored since agreeing
to be an impatient person reflects emotional intolerance, which should obtain
a low score on a scale that measures emotional tolerance.
Q U E S T I O N 4: O P T I O N 3 I S T H E K E Y – A , C A N D D
The scatterplot shows that item 2 has a positive correlation, which means that
the item discriminates well with the total score and between high and low
scores. Therefore, any option with statement B is wrong as the scatterplot does
not show a negative correlation.
...........
54
Lesson
5 5
AIMS
In lesson 5, you will learn why a questionnaire should be reliable and valid
and how to interpret information to determine if a questionnaire meets these
requirements. You will be given activities with examples to teach you how to
apply your knowledge.
Lesson
• correlation coefficient
• reliability
• validity
...........
55 PYC 2616/1
This has serious implications when we try to generalise findings from a study
where the reliability and validity are not as accurate as it was for the original
population. It might lead to inappropriate assumptions about different cultural
groups. This happened during the pre-94 era in South Africa, where intelligence
and aptitude tests that were normed for European populations, were used incor-
rectly on black (and other) South African groups.
5.1 INTRODUCTION
A questionnaire is designed for a specific purpose implying that the researcher
wants to obtain information on a specific concept or construct. But before the
researcher can use a questionnaire, he or she needs to evaluate the functioning
of this questionnaire in terms of two requirements. In the first place, the
following questions need to be answered:
...........
56
...........
57 PYC 2616/1
In summary
Type of reliability Forms of the Coefficient
questionnaire
Administration
sessions
Test-retest reliability One form Coefficient of stability
Two sessions
Alternate-forms Two forms Coefficient of
reliability One session equivalence
(immediate)
Alternate-forms Two forms Coefficient of stability
reliability Two sessions Coefficient of
(delayed) equivalence
Split-half reliability One form Coefficient of internal
One session consistency
...........
58
Insurance companies, for example, need to know that the questionnaires they
use have criterion-related validity. An insurance company might construct a
questionnaire to establish “overall health” in their client population. Does their
current finding of a high or low score in “overall health” say anything about
how much they will claim from the company, or, what their health will look
like a year from now?
...........
59 PYC 2616/1
In summary
Category of validity Example Relevant content/
criterion/comparison
Content validity Questionnaire on Personal experiences of
experience of crime- crime
related stress Reaction to crime
Stress levels
Criterion-related Questionnaire on Managers’ rating
validity organisational commitment of commitment.
Questionnaire on a study Examination
field performance
Construct validity Questionnaire on Community
sociability involvement (high
correlation)
Mechanical interest (low
correlation)
5.4 GLOSSARY
Here are some basic terms related to the topics of reliability and validity that
you may not be familiar with, but which you need to know. They will come
up in your reading of the material in this lesson.
...........
60
...........
61 PYC 2616/1
5 AC TIVITIES
Before you work through the activities, I want you to read the resource
material for this lesson (correlation coefficient, reliability and validity).
T ime re cord
These activities took me ........ hours.
METHOD Once you have compiled the final version of your questionnaire (lesson 4),
there are two more requirements that your questionnaire should meet before
Activity 5.1
you can use it. The first is that the results should be reliable, implying that
Evaluate the
reliability of the the questionnaire should measure consistently. The type of reliability and the
rating scale approximate value of the reliability coefficient need to be considered when
Resource
you evaluate the reliability of your questionnaire.
Correlation
coefficient
Reliability
AC T I O N 01
You should be able to distinguish between different types of reliability. The
purpose of the questionnaire determines which type of reliability is appropriate.
Part of your questionnaire consists of a rating scale aimed at measuring a
specific concept or construct, that is, a specific attribute relevant to the topic
you are investigating. In lesson 4, you analysed the items in this rating scale
and selected those items that are the best measure of this construct for the
final version of your questionnaire. In other words, you want the items in the
rating scale to all measure the same concept or construct. But how do you
know if the items of your rating scale all measure the same thing?
To answer this question, you will check if the rating scale is internally consistent.
Study the resource on reliability and discuss the type of reliability you would
use if you want to determine the internal consistency of a rating scale (the
extent to which the items measure the same thing). Describe the method you
would use and indicate if there are any advantages or limitations to this method.
Because you are interested in the internal consistency of the rating scale in
your questionnaire, you need to obtain an estimate of the split-half reliability
of this rating scale. This will indicate the degree of equivalence between two
halves of the rating scale, that is, the extent to which the two halves measure
the same attribute. As such, it indicates the degree of relatedness of the items
in your rating scale. A limitation of this method is that the reliability coefficient
...........
62
that one obtains to some extent depends on the items included in each of the
two halves. The way in which you divide the rating scale into two equivalent
halves thus influences the value of the reliability coefficient.
AC T I O N 02
You are not expected to calculate a split-half reliability coefficient for the rating
scale in your questionnaire. We will, however, show you how to obtain some
indication of the reliability of either of two halves of the rating scale. This will
enable you to comment on the reliability of this scale.
Below is an example of the responses to the eight items of a five-point rating
scale (i.e. the respondents marked a response from 1 to 5 for each item). Only
five cases are shown here to illustrate the method.
We now have two sets of five scores each (the number of respondents): a set
of total scores on the odd items and a set of total scores on the even items.
The relation between these two sets of scores will give you an estimate of the
reliability of either of the two halves of the rating scale.The relation between
the two sets of scores can be visualised as follows. For each person, we take
the total score on the odd items and the total score on the even items and
where the two meet, we make a dot on a graph (see below). If we do this for
all the people in the group, we will have a scatterplot (a number of scattered
dots). We then draw a straight line resembling the shape of the scatterplot.
...........
63 PYC 2616/1
The upward slope of the line tells you that the correlation coefficient falls
between 0 and +1. Because most dots are close to the line, the correlation
coefficient is close to +1 and there is a fairly strong relation between the two
halves.
You can now repeat this process with your own questionnaire. In lesson 4,
you did an item analysis and selected eight items for the final version of your
questionnaire. Look at these eight items and divide the rating scale into two
halves by grouping the odd items (items 1, 3, 5, etc) together and by grouping the
even items (items 2, 4, 6, etc) together. On your original data sheet (consisting
of 12 items) in lesson 4, draw a line through each of the four items you are
excluding from the final version. Now re-number your items from 1 to 8 before
copying the information to the data sheet below. For each person calculate
the total score for the odd items (add up the scores on the items) and the total
score for the even items.
You now have two sets of 15 scores each (the number of respondents): a set
of total scores on the odd items and a set of total scores on the even items.
For each person, take the total score on the odd items and the total score on
the even items and where the two meet, you make a dot on the graph below.
Do this for all the people in the group to get a scatterplot and then draw a
straight line resembling the shape of the scatterplot.
...........
64
People who frequently use scatterplots can become quite good at estimating
how strong a correlation is represented by a particular scatterplot. Your estimate
is unlikely to be very accurate, but since we do not actually require you to
calculate a correlation coefficient an estimate will have to do.
AC T I O N 03
In this context, the correlation coefficient is a reliability coefficient. Consider
the type of reliability and the approximate value of the reliability coefficient
and evaluate the reliability of the rating scale. Write this down.
You have studied the relation between two halves of the rating scale and
therefore have an indication of the reliability of either of the two halves. This
indicates the extent to which the halves are equivalent and thus measure the
same attribute. Based on your scatterplot and the straight line you drew, you
“guessed” the value of your reliability coefficient. The values can range from 0
to 1 with values closer to 1 indicating a more reliable rating scale. For example,
if the reliability coefficient for the rating scale on levels of stress associated
with different crimes (the example used in lessons 1 and 2) was 0.85, this
would imply that the items in this scale to a large extent measure the same
attribute, namely stress.
Note that you only have the reliability of one half of the rating scale but the
split-half reliability of the whole rating scale would probably be even higher
as a shorter questionnaire is generally less reliable than a longer questionnaire.
...........
65 PYC 2616/1
Activity 5.2 The interpretations based on the results of your questionnaire should not only
Evaluate the validity be reliable but should also be valid, that is, the questionnaire should measure
of the questionnaire
what it claims to measure. A questionnaire is valid for a specific purpose and
Resource you should gather evidence relevant to the purpose of your questionnaire to
Correlation determine its validity.
coefficient
Validity
AC T I O N 01
You should be able to distinguish between categories of validity. In lesson 1,
you stated the purpose of your questionnaire and based on this you identified
the content domain of your questionnaire. In lesson 2, you designed your
questionnaire so that the items are representative of the specified content
domain. The degree to which you succeeded in this, influences the validity
of your questionnaire. Study the resource on validity and discuss the category
of validity that is relevant to questionnaire design. Write this down so that you
can formulate your ideas clearly.
AC T I O N 02
Consider the content domain of your questionnaire (lesson 1) and the
questionnaire specification document (lesson 2) and evaluate the content
validity of your questionnaire. Write down your thoughts on this before you
read the next paragraph!
It should be clear from your definition of the content domain (that is the relevant
universe of tasks, behaviours, attitudes) whether any particular task, behaviour
or attitude forms part of this universe. The content of the items and the item
format should indeed measure the tasks, behaviours or attitudes you defined.
The number of items should furthermore ensure that the universe of all tasks,
behaviours or attitudes is adequately represented. In lesson 2, an example of
a questionnaire for an investigation of South Africans’ experience of crime-
related stress was discussed. The content areas, type of items, and number of
items were in accordance with the purpose and the content domain of this
questionnaire as identified in lesson 1.
...........
66
S E L F -A SS E SS M E N T
Study the scenario carefully before you consider the self-assessment questions.
The leisure activities questionnaire (LAQ) can be used for adults to obtain
information on what people do during their free time (that is, when they are
not busy with compulsory activities). The questionnaire was administered to
a representative sample of adults during the pilot-testing to determine the
reliability and validity of the questionnaire. The respondents were tested on
two occasions with two weeks between the testing occasions. The scores of 5
of the respondents for the first and second applications of the LAQ are given
in table 1.
Table 1: Five respondents’ test scores for two administrations of the LAQ
Respondent 1 2 3 4 5
First application 50 43 60 49 56
Second application 48 32 38 55 60
The respondents also completed two additional questionnaires, the sport
participation questionnaire (SPQ) and the numerical skills questionnaire
(NSQ). The results of these measurements are shown in table 2. It contains
the correlations among the questionnaires.
Table 2: Correlations between the LAQ and other questionnaires
LAQ SPQ NSQ
LAQ 1.0
SPQ 0.84 1.0
NSQ 0.12 0.18 1.0
QUESTION 1
Lerato: What does the information in table 1 tells us about the LAQ?
Ben: Actually, the information in table 1 does not reflect a poor correlation,
which means that the LAQ shows good criterion-related validity.
...........
67 PYC 2616/1
You agree most with ..., second most with and least with
QUESTION 2
A year programme has been introduced in a school to make the children
more aware of environmental conservation. The children will complete a
questionnaire to measure their attitude to and knowledge on issues related
to environmental awareness. This questionnaire is administered in January
and again in July and in December. Which type of reliability do you think
is appropriate and how would you interpret a value of 0.60 on this type of
reliability?
QUESTION 3
In table 2, what does the correlation between the LAQ and SPQ mean? Choose
the most correct answer.
(1) It means there is a poor relationship between the LAQ and SPQ.
(2) It means the results on the two questionnaires are related.
(3) It shows that the LAQ has concurrent validity.
(4) It shows that the LAQ has discriminant validity.
QUESTION 4
In table 2, what does the correlation between the LAQ and NSQ mean? Choose
the most correct answer.
(1) It means there is a poor relationship between the LAQ and NSQ.
(2) It means the results on the two questionnaires are related.
(3) It shows that the LAQ has concurrent validity.
(4) It shows that the LAQ has discriminant validity.
QUESTION 5
Define the content domain for the following two questionnaires: a questionnaire
to measure spelling achievement in Grade 2 and a questionnaire to measure
anxiety.
...........
68
Q U E S T I O N 1: O P T I O N 1 I S T H E K E Y – S U S A N , V U S I A N D B E N
Order the values on the first administration in ascending order:
Person 2 4 1 5 3
1st 43 49 50 56 60
Now write down the corresponding values for the second administration:
Person 2 4 1 5 3
1st 43 49 50 56 60
2nd 32 55 48 60 38
Can you see that there is little correspondence between the pattern of values
on the first administration (ascending) and that of the second administration
(a mix of high and low values with no clear pattern)? This implies that there
is a poor correlation between the two sets of scores. A correlation between
two administrations of the same questionnaire at different times refers to the
test-retest reliability of a questionnaire. Susan was therefore correct on both
issues. Vusi was correct in identifying that the correlation is poor but this does
not refer to criterion-related validity. Ben was incorrect on both issues.
Q U E S T I O N 2:
You would have been correct if you indicated that the test-retest reliability
should be examined. This questionnaire is administered to the children more
than once and should thus demonstrate consistency over time. A value of 0.60
indicates that differences in results can be expected over different occasions
(remember the values of the reliability coefficient range from 0 to 1). The value
is nevertheless not so low that it will not provide any information but we would
suggest that the school uses the results in combination with other information.
Q U E S T I O N 3: O P T I O N 3 I S T H E K E Y
We would expect the results on the leisure questionnaire and the sports
questionnaire to be related. A value of 0.84 indicates a good correlation
(i.e. a relationship) between the LAQ and the SPQ meaning that option 2 is
correct. One of the methods to determine construct validity is to consider the
concurrent validity (the relationship between your questionnaire and another
questionnaire that measures a similar construct) and discriminant validity (a
low correlation between your questionnaire and another questionnaire that
measures a different construct). The correlation between the LAQ and the SPQ
therefore shows concurrent validity making option 3 the most correct answer.
Q U E S T I O N 4: O P T I O N 4 I S T H E K E Y
We would not expect a relationship between the results on the leisure
questionnaire and the numerical skills questionnaire. A value of 0.12 indicates
...........
69 PYC 2616/1
a low correlation (i.e. the absence of a relationship) between the LAQ and
the NSQ meaning that option 1 is correct. The correlation between the LAQ
and the SPQ therefore shows discriminant validity, making option 4 the most
correct answer.
Q U E S T I O N 5:
You will find it easier to define the universe of material and tasks for measuring
achievement where a great deal is known about that which you want to
measure. You can compile a list of nearly all possible words that a Grade 2
child should be able to spell. You will then select words ranging in difficulty
level from this list. You should also decide on the tasks, for example, the items
could be multiple-choice items. Content validity is more difficult to assure
when a questionnaire measures a more abstract attribute. It is not possible,
for example, to provide all possible items for a measure of anxiety. You would
have to rely on the opinion of experts as to whether the items represent the
content domain.
...........
70
Lesson
6 6
6 Compile a manual
AIMS
In lesson 6, you will learn why it is important to have a manual for a questionnaire,
how to logically structure your manual, what information about the question-
naire should be provided and in how much detail. You will be given activities
with examples to teach you how to apply your knowledge.
Lesson
...........
71 PYC 2616/1
the purpose of the questionnaire, the possible importance for the community
we are engaging with should be mentioned.
The description of the target population should refer to the cultural and lan-
guage group/s for which the questionnaire is suitable. In a community project,
one would involve the various stakeholders to ensure that the community (the
target population) benefits as much from the co-designed questionnaire as the
researcher does.
6.1 INTRODUCTION
A researcher identifies the topic for his or her study based on a number of
considerations including perceived community concerns. The researcher
might decide to use an existing questionnaire in the study but he or she needs
information on this questionnaire to decide if it is suitable for the study. We
will call the researcher the user of the questionnaire and the purpose of the
manual is to provide the user with information on the following:
Can you see that the manual is a summary of all the work that you did in
lessons 1 to 5?
6.2.3 Does the design of the questionnaire ensure that it meets its aim?
A brief description of the design of the questionnaire is provided in the manual.
This includes the domain of the questionnaire and an indication of how the
items cover this domain. The item types are specified and it is indicated how
a specific item format is suitable to obtain specific information. The layout of
the questionnaire is also mentioned.
...........
73 PYC 2616/1
Do you see that you might have a total score for more than one section of the
questionnaire? Guidelines for interpretation should then be provided for each
of these sections. In the case of right or wrong answers, a total score might
indicate the degree of knowledge on a topic and someone with a high score
knows more than someone with a low score. A total score on a rating scale
usually indicates an attitude towards a certain issue and one can distinguish
between individuals in terms of their attitude. You can therefore see that the
guidelines for the interpretation of the results are based on the aim of the
questionnaire.
In summary
6.5 GLOSSARY
Here are some basic terms related to the manual that you may not be familiar
with, but which you need to know. They will come up in your reading of the
material in this lesson.
...........
75 PYC 2616/1
6 AC TIVITIES
Before you work through the activities, I want you to read the resource
material for this lesson (manual: purpose and structure).
Manual: purpose
and structure
Press the button
or go to page 132 in
the study guide.
...........
76
T ime re cord
These activities took me ........ hours.
METHOD In the previous lessons, you went through a process of developing and evaluating
a questionnaire. You now need to compile a manual in which this process
Activity 6.1
is described so that any potential user can know when and how to use
Discuss the process
of developing the the questionnaire and also have the information to determine how well the
questionnaire questionnaire will work for the intended purpose. Generally, such a manual
consists of a description of the aim and design of the questionnaire, an evaluation
Resource
of its properties, and the procedures required for administration, scoring and
Manual: purpose
and structure interpretation.
AC T I O N 01
In lesson 1, you defined the purpose of your questionnaire: you stated the aim
of the questionnaire (what it measures and how this information can be used)
and you indicated the target population (for whom it is intended). Write down
the aim statement of your questionnaire and describe the target population of
your questionnaire, clearly indicating which characteristics are important. Refer
to the suitability of the questionnaire in a multicultural and multilingual context.
In lesson 2, you described your items in terms of content and format. You
also specified the layout of the questionnaires. Use this information to give a
brief description of how the design of the questionnaire ensures that it meets
its purpose.
AC T I O N 02
In lesson 4, you administered your draft questionnaire to a group of people
who were representative of your target population. You used their results to
analyse the items and to determine the spread of responses for each item and
the relation between each item and the total over all the items. Briefly describe
how you used each of these techniques to select items for the final version of
your questionnaire. Provide an example of an item that you included and an
item that you excluded. Illustrate this with a scatterplot to show both the spread
of responses and the relation with the total score for each of the two items.
AC T I O N 03
You have written instructions for the administration of the questionnaire in
lesson 2. Write down your introductory talk and instructions for completing the
questionnaire. Check that you have referred to the following: the administrator;
the situation; the material; dealing with questions; the aim and importance;
and confidentiality.
In lesson 2, you also selected a scaling method. This can be used to assign
numerical values to the answers on the items and to score the rating scale. In
lesson 4, you had to score the rating scale in your questionnaire. In the manual,
you have to provide instructions for scoring those items in the questionnaire
that require scoring. Indicate how each item is scored and how item scores
are summed to get a total score. Write this down.
Lastly, you need to provide guidelines for the interpretation of the results of
the questionnaire. You should indicate whether the interpretation is based
on the response to each item separately, whether (as in the case of the rating
scale) the total score for a section is interpreted or whether the total score for
the questionnaire is used. Remember that the interpretation is linked to the
aim of the questionnaire (what the questionnaire measures).
Activity 6.2 The information you have is sufficient to compile a mini-manual for your
Compile a manual questionnaire. Copy the material from the previous exercises (activity 6.1)
under the following headings.
Resource
Manual: purpose AIM AND DESIGN
and structure
S E L F -A SS E SS M E N T
Study the scenario carefully before you consider the self-assessment questions.
You are a human resources (HR) officer and your company requested you
to determine how well the personnel are coping with work-related stress.
You decide to administer a number of questionnaires and one of these is the
leisure activities questionnaire (LAQ). Below is a section from the manual for
the questionnaire.
Mini manual for the leisure activities questionnaire (LAQ)
1. Aim and design
The leisure activities questionnaire (LAQ) can be used to obtain information
on what people do during their free time (that is, when they are not busy
with compulsory activities). Recreational or leisure activities are important to
ensure a balanced life and to help people cope with stress and it is important
to know if people are in fact partaking in these types of activities. The LAQ
can be used for any adult regardless of gender, education level, population
group, etcetera. There are three factors that are important when studying
leisure activities, namely the type of activity (whether it is a sport, a cultural
or a social activity), the consistency with which the activity is practiced (how
often and for how long) and the reason for doing the activity (for enjoyment,
amusement, etcetera). The LAQ consist of 60 items, 20 for each factor, that have
been developed together with representatives from various communities. This
was done because one would expect that socio-cultural factors would have
an influence on the leisure activities that people partake in and the various
possibilities had to be covered in the questionnaire. A mixture of inventories,
rating items and open questions were included.
QUESTION 1
Susan: I presume a good place to begin this discussion is the purpose of
the LAQ.
Lerato: Well, the purpose of the questionnaire clearly is to find out more
about the leisure activities that people partake in.
...........
79 PYC 2616/1
You agree most with, second most with and least with
QUESTION 2
The target population for the LAQ is adequately described.
QUESTION 3
Which of the following is NOT a purpose of a manual?
Q U E S T I O N 1: O P T I O N 1 I S T H E K E Y – L E R ATO, B E N A N D V U S I
The purpose of the LAQ is to find out more about the type of activities that
people partake in, the consistency with which they practice the activities and
their reasons for doing the activities. That is, to find out more about the leisure
activities that people partake in. The intention of the HR officer is to find out
how well people are coping with work-related stress – he/she can use the LAQ
to get more information on the topic but the LAQ could also be used for other
types of projects as its main purpose is not to determine coping. The statement
that the purpose is to assess people in the workplace could be made about
many questionnaires and is not specific to the purpose of the LAQ.
Q U E S T I O N 2: O P T I O N 2 I S T H E K E Y
The target population is described in detail (age, gender, educational level and
population group). Although age (option 1) and gender (option 3) are indeed
specified, each on its own would not be sufficient to say that the description
of the target population is adequate. Provision was made for differences in
culture during the design and option 4 is therefore incorrect. The only aspect
that remains unclear is whether the functioning of the questionnaire has been
determined for different language groups (option 2).
...........
80
Q U E S T I O N 3: O P T I O N 3 I S T H E K E Y
Although the process followed to select items is discussed in the section dealing
with the properties of the questionnaire, the items have already been selected.
That is, the information is not provided to enable item selection.
...........
81 PYC 2616/1
Lesson
7 7
7 Review a questionnaire
In this lesson you will use everything you have learned in lessons 1, 2 and 3
to review a questionnaire.
AIMS
In lesson 7, you will learn how to use a rating scale to evaluate the extent to which
each aspect of a questionnaire meet the requirements in terms of correctness,
suitability and functionality. You will be given activities with examples to teach
you how to apply your knowledge.
Lesson
• content domain
• item format
• layout of the questionnaire
• suitability of the questionnaire
• writing questionnaire items
the value is to them and their community. The items of the questionnaire should
be formulated in such a way that it is clear to all the respondents and it should
also be suitable to diverse contexts. The questionnaire will be functional for
different groups if the information on the intended content domain is gathered
for all respondents included in the target population.
7.1 INTRODUCTION
In lessons 1, 2 and 3, you designed and developed your own questionnaire.
You identified a suitable content domain, your questionnaire specification
document indicated the type of items, number of items, layout and so on
of your questionnaire, and you followed certain guidelines in writing items
for your questionnaire. In this lesson, you will apply the competencies you
acquired in developing your own questionnaire to evaluate a questionnaire
developed by someone else.
To evaluate a questionnaire you will use a rating scale that consists of four
sections. These sections correspond to the main aspects of a questionnaire. The
four main aspects one should consider when evaluating a questionnaire are
Thus, there are three points to consider when a questionnaire’s instructions are
reviewed: the purpose of the questionnaire, the issue of confidentiality, and
the instructions for completing the questionnaire. The following rating item is
used to evaluate the instructions of a questionnaire:
...........
84
(a) the respondents are likely to find the language level appropriate
(b) the respondents are likely to find the item relevant in their context
(c) the respondents are likely to possess the knowledge required to respond
correctly
You can read more about these topics in the resources dealing with the
identification of a questionnaire’s content domain, the layout of a questionnaire,
and the writing of questionnaire items. The following rating item is used to
judge the structure of a questionnaire:
...........
85 PYC 2616/1
...........
86
A questionnaire is functional if
7.6 IN SUMMARY
...........
87 PYC 2616/1
7.7 GLOSSARY
Here are some basic terms related to the review of a questionnaire that you
may not be familiar with, but which you need to know. They will come up in
your reading of the material in this lesson.
Sephiri
Anonymity A person’s name/ ’n Persoon se Ha ho sa tsejwe
Anonimiteit identity is unknown. naam/identiteit is lebitso la motho.
onbekend.
Ho se bolele lebitso
Functionality Being suited to serve Om gepas te wees om -loketseng ho fihlella
Funksionaliteit a purpose. ’n doel te dien. sepheo.
Ho kgona ho etsa
mosebetsi
7 AC TIVITIES
Before you work through the activities, I want you to read the resource
material for this lesson (content domain, item format, layout of the
questionnaire, suitability of the questionnaire and writing questionnaire
items).
Suitability of a Writing
questionaire questionnaire items
Press the button Press the button
or go to page 141 in or go to page 147 in
the study guide. the study guide.
...........
88
T ime re cord
These activities took me ........ hours.
METHOD The questionnaire rating scale consists of the five rating items discussed in the
introduction. These are the criteria required when you review a questionnaire.
Activity 7.1
You will now get the opportunity to apply these criteria.
Use the
questionnaire rating
scale to review a
questionnaire AC T I O N 01
Resource Study the following questionnaire:
Content domain
Layout of a SIGNIFICANT ASPECTS OF RESIDENTIAL
questionnaire
Writing AREAS QUESTIONNAIRE
questionnaire items
The significant aspects of residential areas questionnaire (SARAQ) measures
what individuals consider to be important aspects of their residential areas.
The questionnaire is intended for individuals of age 18 years and older. The
respondents should have been resident in the area for at least six months
before completing the questionnaire.
For the purpose of this questionnaire, residential areas are viewed in terms of
high density, medium density and low density areas. High density areas refer
to living units in blocks of multi-storey buildings (i.e. buildings consisting of
three or more floors). Medium density living areas contain duplex and simplex
living units. Low density living areas contain single freestanding homes. Living
density plays an important role in what people consider to be significant aspects
of their living areas. The primary aspects of any living environment are issues
concerning (a) necessity/desire, (b) comfort/ convenience, (c) recreation, and
(d) security.
Read each statement carefully and tick the relevant box for your answer. (SA
= Strongly agree, A = Agree, D = Disagree, and SD = strongly disagree)
SA A D SD
1 It is more important to be warm in winter than
to be close to the supermarket.
2 A house should be proper.
3 I want a beautiful house and a happy family.
4 I hate to spend more money on my electricity
bill than on my golfing fees.
5 Every home should have a television set.
6 When did you move into the house you are
staying in at present?
...........
89 PYC 2616/1
SA A D SD
AC T I O N 02
Refer to the rating scale discussed in the introduction and use the following
review sheet to review the questionnaire. Tick the number that corresponds
to the rating you assign.
...........
90
Activity 7.2 Your ratings may differ from the standard set of ratings that are provided here.
Compare your In most cases you will probably agree that the answers given are better. Here
reviews to a
standard set of are the questionnaire’s standard set of ratings. The explanations for the ratings
ratings follow below.
1. Rating: 2
2. Rating: 4
...........
91 PYC 2616/1
3. Rating: 4
4. Rating: 2
5. Rating: 4
Item 2 is suitable. The respondents are likely to find the language level
appropriate and the item relevant in their context. They are also likely to
possess the knowledge required to respond.
6. Rating: 3
7. Rating: 4
Item 3 is suitable. The respondents are likely to find the language level
appropriate and the item relevant in their context. They are also likely to
possess the knowledge required to respond.
8. Rating: 4
9. Rating: 2
Item 4 may not be suitable. The respondents are likely to find the lan-
guage level appropriate. But they may not find it relevant. The item is
relevant only to those who play golf. People who do not know what the
going rate for golfing fees is do not possess the knowledge required to
respond correctly.
...........
92
10. Rating: 4
11. Rating: 2
Item 5 may not be suitable. The respondents are likely to find the language
level appropriate. But they may not find it relevant. The item is relevant
only to those who can afford a television set. However, they are likely
to possess the knowledge required to respond correctly.
12. Rating: 1
13. Rating: 4
14. Rating: 4
15. Rating: 1
Item 7 is not suitable. Not all the respondents are likely to find the lan-
guage level appropriate, or to find the statement to be relevant in their
context (especially respondents in lower socio-economic groups who are
more likely to rent than to buy property). Furthermore, these individuals
will probably not have the knowledge required to respond correctly.
16. Rating: 4
...........
93 PYC 2616/1
17. Rating: 2
Item 8 may not be suitable. The respondents are likely to find the language
level appropriate. But they may not find it relevant. The item is relevant
only to those who can afford a car. However, they are likely to possess
the knowledge required to respond correctly.
18. Rating: 1
19. Rating: 2
Item 9 is not suitable. The respondents are likely to find the language ap-
propriate but individuals from lower socio-economic circumstances may
not find the idea of a holiday away from home relevant in their context,
although they do have the knowledge to respond to the question.
20. Rating: 2
Item 10 is not formulated correctly. The item has content and construct
relevance (it refers to convenience), it is configured correctly, is not
leading and not likely to elicit a socially desirable response. However,
it is difficult to comprehend because it is formulated in an unclear and
ambiguous manner.
21. Rating: 4
Item 10 is suitable. The respondents are likely to find the language level
appropriate, and the question relevant. They are likely to possess the
knowledge to answer the question correctly.
22. Rating: 3
The questionnaire has sufficient scope in the sense that it refers to all four
content/construct areas. It does not encourage inefficient responding (it
requires all items to be responded to, and the items can be completed
in the order in which they are presented). It is not insensitive towards
respondents as there are no items that may cause embarrassment. How-
ever, the questionnaire encourages a particular response style. The re-
spondents are likely to agree with most statements, leading to the left or
centre columns to be marked more often.
23. Rating: 2
...........
94
S E L F -A SS E SS M E N T
Activities 7.1 and 7.2 serve as self-assessment. We will, however, repeat two
of these questions to show you how we will usually ask the questions.
QUESTION 1
Rate the formulation of item 1.
QUESTION 2
Rate the suitability of item 4.
Item 4 is suitable if
(a) the respondents are likely to find the language level appropriate
(b) the respondents are likely to find the item relevant in their context
(c) the respondents are likely to possess the knowledge required to respond
correctly
Q U E S T I O N 1: O P T I O N 4 I S T H E K E Y
See the explanation in activity 7.2.
Q U E S T I O N 2: O P T I O N 2 I S T H E K E Y
See the explanation in activity 7.2.
...........
95 PYC 2616/1
Lesson
8 8
8 Review a manual
In this lesson you will use everything you have learned in lessons 1 to 6 to
review a manual. The information in lesson 6 is especially important.
AIMS
In lesson 8, you will learn how to use a rating scale to evaluate the extent to which
a manual covers the required content, is logically structured and is written in
clear and comprehensible language. You will be given activities with examples
to teach you how to apply your knowledge.
Lesson
...........
96
diverse contexts. And the sociocultural context should be accounted for in the
instructions for interpretation.
8.1 INTRODUCTION
In lesson 6 you compiled a manual in which the process of developing and
evaluating your questionnaire was described. In this lesson you will apply
the competencies you acquired in writing a manual for your questionnaire to
evaluate a manual written by someone else.
To evaluate a manual, you will use a rating scale that consists of three sections.
These sections correspond to the main criteria for evaluating a manual. The
three main criteria one should consider when evaluating a manual are
...........
97 PYC 2616/1
...........
98
...........
99 PYC 2616/1
In summary
Purpose (a) the aim of the questionnaire is stated
(b) the target population is defined
(c) the design of the questionnaire is described
Properties (a)
the sample used to test the questionnaire is described
correctly
(b)
item analysis and selection procedures are described
correctly
(c)
the reliability of the questionnaire is discussed correctly
(d) the validity of the questionnaire is discussed correctly
Procedures (a)
instructions are provided for administering the
questionnaire
(b)
instructions are provided for scoring the questionnaire
(c) instructions are provided for interpreting results
Structure (a) the text is clustered into topical sections
(b) the topics are presented in logical order
(c)
the information about each topic is relevant and presented
coherently
Language (a) the text is factual and objective
(b)
the text is clear and to the point, with no ambiguous and/
or conflicting statements
(c)
the text is not inundated with numerous spelling, grammar
and punctuation errors
8.5 GLOSSARY
Here are some basic terms related to the topics of reliability and validity that
you may not be familiar with, but which you need to know. They will come
up in your reading of the material in this lesson.
-sehlooho
Factual Based on facts Gebaseer op feite. -itshetlehileng ka
Feitelik dintlha tsa nnete
-nnete
Ambiguous Open to more than Oop vir meer as een -hlalosehang ka
Dubbelsinnig one interpretation. interpretasie. ditsela tse ngata.
-sa hlakang
...........
100
-utlwisisehang
Coherent Logical and Logies en goed -utlwisisehang le ho
Samehangend well-organised. georganiseerd. hlophiswa hantle.
Utlwisiseha
8 AC TIVITIES
Before you work through the activities, I want you to read the resource
material for this lesson (manual: purpose and structure).
Manual: purpose
and structure
Press the button
or go to page 132 in
the study guide.
T ime re cord
These activities took me ........ hours.
METHOD The manual rating scale consists of the five rating items discussed in the
introduction. These are the criteria that are required when you review a manual.
Activity 8.1
You will now get the opportunity to apply these criteria.
Use the manual
rating scale to review
a manual
Resource AC T I O N 01
Manual: purpose Study the manual below.
and structure
MARSTON PERSONALITY QUESTIONNAIRE MANUAL
...........
102
The questionnaire consists of 60 multiple choice items. The four factors are
covered in each item. Each item has the following format:
If [description of work situation] I prefer to:
(a) [description of ‘drive’ action]
(b) [description of ‘interaction’ action]
(c) [description of ‘management’ action]
(d) [description of ‘regulation’ action]
For example:
If things get really hectic at work I prefer to:
(a) put my mind to it and get things done as quickly as possible
(b) consult with my colleagues
(c) dig in and maintain a steady work pace
(d) stick with company policy
To counter response biasse the sequence in which the action descriptions are
provided is varied randomly. For example, the [description of ‘drive’ action]
is not always placed first. Fifty different [descriptions of work situations] are
used to cover a wide range of work events.
A large sample of nearly nine hundred university students was used in the
development of the questionnaire. The sample came from a number of different
universities across the country and was representative of the south african
population in respects of gender, race and education levels. All students were
in their final year of study and ready to enter the job market. The original MPQ
that was used in the sample study consisted of 145 items. An item analysis
showed that 67 of these items were really good but that the remaining 78 items
did not meet the criteria to be included. A further seven of the 67 selected items
were excluded because they overlapped with [description of work situation].
Furthermore 60 items were retained for the final version of the MPQ. The
large number of good items ensures a high degree of questionnaire reliability,
and supports the questionnaire’s construct validity. The validity coefficient of
0,91 is very high for this kind of questionnaire.
– End of manual –
AC T I O N 02
Refer to the rating scale discussed in the introduction and use the following
review sheet to assess the manual. Tick the number that corresponds to the
rating you assign.
...........
103 PYC 2616/1
Activity 8.2 Your ratings may differ from the standard set of ratings that are provided here.
Compare your In most cases, you will probably agree that the answers given are better. Here
reviews to a standard are the manual’s standard set of ratings. The explanations for the ratings follow
set of ratings
below.
1. Rating: 4
The aim of the questionnaire is stated: The manual indicates that the MPQ
was designed to measure individuals’ preferred behaviour styles in their work
environments.
The target population is defined: According to the manual the MPQ can be used
for any working person, that it is applicable from the most basic to the most
sophisticated jobs, and that it works across gender, race and education levels.
2. Rating: 1
The manual does poorly in this regard. It meets none of the review criteria.
The sample used to test the questionnaire is not described correctly if one
views this description in light of the questionnaire’s target population. The
questionnaire’s target population is working people but the sample used to
test the questionnaire comes from a student population.
The manual mentions item analysis and selection but it does not describe
how this was done. We read that 78 items did not meet the selection criteria
but we do not know what these criteria were. Only one selection criterion is
mentioned, namely that seven items were deleted due to their overlap with
[description of work situation]. However, this has to be incorrect. All items
should have a [description of work situation] otherwise the questionnaire will
not have face and content validity. Therefore this statement may be a misprint
(perhaps the intention was to say that the items were rejected because of poor
or low overlap), or else the author of the manual simply did not know what
he/she was talking about.
...........
104
Validity and reliability: The manual mentions validity and reliability but there is
no proper description and discussion of these properties of the questionnaire.
3. Rating: 4
4. Rating: 1
The manual is not structured properly. The text is not clustered into topical
sections, topics do not follow on each other in logical order, and although the
information about each topic is relevant, it is not always presented coherently.
...........
105 PYC 2616/1
Clustering into topical sections: The manual does not use any headings to
organise its content. The text should have been structured under headings
such as:
Each of these topical sections could have been structured further in terms of
subsections. For example:
The logical order of topics: The manual does not present the topical sections
in logical order. The first two paragraphs deal with the administration of the
MPQ. The purpose of the questionnaire, which should have been indicated
first, is only introduced in the third paragraph. The rest of the text is structured
logically though. The questionnaire is described in terms of its items, and this
is followed by an explanation of how the questionnaire was pilot-tested.
...........
106
...........
107 PYC 2616/1
5. Rating: 3
The manual is written in a factual and objective style, and the text is clear and
to the point, with no ambiguous and/or conflicting statements. Unfortunately,
the manual contains numerous spelling, grammar and punctuation errors.
Incorrect Correct
First paragraph:
instrucsions instructions
A pencil are required A pencil is required
an answer sheet. Which can be an answer sheet, which can be
questionnaire were been completed questionnaire should be completed
Second paragraph:
to allow his/her to ask to allow him/her
Each answer count Each answer counts
The 4 scale values The four scale values
behavior behaviour
interprate interpret
Third paragraph:
All person has Every person has
People who does not have People who do not have
Fourth paragraph:
drive {the ability drive (the ability
abylity ability
...........
108
Seventh paragraph:
biasse bias
Eight paragraph:
the south african the South African
Furthermore 60 items Thus 60 items
S E L F -A SS E SS M E N T
Activities 8.1 and 8.2 serve as self-assessment. We will, however, repeat two
of these questions to show you how we will usually ask the questions.
QUESTION 1
Rate the manual’s indication of the purpose of the MPQ.
QUESTION 2
Rate the structure of the manual.
...........
109 PYC 2616/1
Q U E S T I O N 1: O P T I O N 4 I S T H E K E Y
See the explanation in activity 8.2.
Q U E S T I O N 2: O P T I O N 1 I S T H E K E Y
See the explanation in activity 8.2.
...........
110
RESOURCES
Content domain
Correlation coefficient
Item analysis
Item format
Layout of a questionnaire
Manual: purpose and structure
Reliability
Specification document
Suitability of a questionnaire
Validity
Writing questionnaire items
...........
111 PYC 2616/1
Content domain
The first step is to identify the general topic of interest. For example, you may
be interested in the interface between people and technology, or in the topic of
crime, or people’s shopping habits. These are very broad topics, so you need to
narrow them down further before you can investigate them properly. The next
step is to select a problem area within that topic or one aspect of the topic that
you want to investigate. For example, if you are interested in the topic of crime,
you should select one aspect of crime, such as the impact of crime on people’s
everyday lives – that will give some focus to your investigation. However, the
problem area is still fairly broad, so now you need to reduce the general problem
to more specific questions. To continue with the example of crime, you could
reduce the general problem (crime) to more specific questions like: “What are
people’s perceptions of crime in South Africa?”, “How do people cope with
the level of crime in society?”, “Do South Africans experience crime-related
stress?” or “What are people’s perceptions of the criminal justice system?” All
these questions are relevant to the problem area, namely the impact of crime
on people’s lives, and relate to the broader topic of crime. Having identified
these questions, you have established the focus of your questionnaire. The
content domain therefore consists of the tasks, behaviours, attitudes, etcetera
related to one or more of these questions.
...........
112
The guidelines are that you have to ask (1) enough questions to cover your
content domain adequately but (2) only questions that are directly relevant to
your content domain. Let us look at an example by taking one of the questions
posed in the previous section: “Do South Africans experience crime-related
stress?” This suggests that your questionnaire should include items or questions
dealing with how people feel about crime, whether the crime situation is
stressful for them and so on. Even though the question “Do South Africans
experience crime-related stress” is fairly specific, there are still many other
aspects that relate to it. For example, the experience of crime-related stress
may vary according to people’s personalities or mental health, or the kind of
news reporting they are exposed to, or their cultural belief systems. They might
also be experiencing stress that is caused by other factors apart from crime.
You cannot possibly ask questions that cover all these aspects adequately
and still have a useful questionnaire (see also “specification document for a
questionnaire”). So your task is to limit the scope of your questionnaire, that
is, decide what is directly relevant to the focus of your questionnaire. If the
questions do not seem relevant, your respondents might not want to complete
your questionnaire.
Here are some ideas about what you might include in your questionnaire
to investigate whether South Africans experience crime-related stress. For
example, you could ask:
Once you have made a list of the aspects you want to cover, look at each
one and ask yourself: “Is this directly relevant” and “why?”. This forces you to
have a good reason why each one is included in the questionnaire. By limiting
the scope in this way, you can cover your topic adequately but still have a
questionnaire that is relevant and not too long.
...........
113 PYC 2616/1
Correlation coefficient
2. The scatterplot
If you visualise the relation between two constructs, you get a clear indication of
the strength and direction of this relation. The graphic display of the correlation
coefficient is called a scatterplot. The graph consists of a horizontal line
(called a horizontal axis) and a vertical line (or vertical axis) that meet at an
angle of 90 degrees. Possible scores on the one construct appear on the one
axis and possible scores of the second construct on the other axis. Note that
...........
114
these scores are all the possible scores a person can obtain on a questionnaire
measuring this construct. For each individual you take his or her score on the
one construct and his or her score on the other construct and where the two
meet you make a dot. Because you have a group of people, you will have a
number of scattered dots (each representing a person) and therefore the name
scatterplot. If there is a perfect positive relation between two constructs (a
correlation coefficient of +1), the dots form a perfectly straight line with an
upward slope. This is shown in scatterplot a in figure 1. For a perfect negative
relation (a correlation coefficient of –1), the scores form a perfectly straight line
with a downward slope (see scatterplot b in figure 1). No relation (a correlation
coefficient of 0) between two constructs results in an undefined shape (see
scatterplot c in figure 1).
Scatterplot a
Scatterplot b
...........
115 PYC 2616/1
R E S O U R CE S
Scatterplot c
...........
116
R e s o u rce s
Figure 3: Linear graph of the relation between emotional sensitivity and anxiety
...........
117 PYC 2616/1
R E S O U R CE S
Item analysis
1. Item difficulty/variance
Imagine a questionnaire where everybody always gets all the right answers
or, a questionnaire where nobody ever gets any right answer. It is clear that
such a questionnaire is of no use in showing up differences between people.
Starting from this extreme example, test constructors have argued that the
ideal questionnaire would be one where about half the people who do
the questionnaire gets each of the items on the questionnaire right. Such a
questionnaire would be ideal for identifying differences between people. Item
analysis therefore involves discarding items that are too easy (almost everybody
gets them right) or too difficult (almost nobody gets them right) and keeping
items that are of medium difficulty. The difficulty index for an item is usually
calculated by dividing the number of people who gave a correct response by
the total number of people in the sample. Ideally, the difficulty index for items
in a questionnaire should be between 0.25 and 0.75 (i.e. between 25% and
75% of people get the item right) and the average difficulty index for items in
a questionnaire should be about 0.5.
When dealing with instruments where there are no right or wrong answers
– such as personality questionnaires and attitude scales – we cannot use the
concept of item difficulty in the same way, but the same principle applies.
We do not want items where everybody answers in the same way as they do
not help us to see differences between people. In this module, we will refer
to this kind of item analysis simply as item variance.
2. Item discrimination
This refers to the ability of an item to discriminate between respondents
according to whatever the measuring instrument as a whole is measuring. Items
should only be selected for the final version of the instrument if they measure
the same characteristics as the other items in the instrument – else, they cause
the instrument to lose focus so that it becomes difficult to know what it might
be measuring. Discrimination is measured by correlating each item in a scale
with the total score for the scale (see “correlation coefficient”). The higher
the correlation coefficient, the more discriminating the item. A minimum
...........
118
R e s o u rce s
correlation of 0.2 is generally required, but this can vary depending on what
kind of construct is being measured. Items with negative or zero correlations
are almost always excluded. A negative correlation could be indicative that an
item should have been reverse-scored, and before discarding such an item, it
is important to check if this is the case.
...........
119 PYC 2616/1
R E S O U R CE S
Item format
1. Closed questions
Generally speaking, most questions are either “open” or “closed”. A closed
question is one that offers respondents a limited choice of alternate replies
whereas an open question allows the respondents to answer in any way
they want to. Examples of closed questions would be those that offer simple
choices, such as yes or no, or a range of alternatives, such as five different
brand names. They are considered “closed” because the respondents do not
have the freedom to answer any way they want to.
Closed questions of the yes/no type merely require respondents to tick off the
answer “yes” or “no”, according to whether they agree with the statement or
not. Here is an example of a closed question of the yes/no type:
Closed questions, the true/false type, are similar to the yes/no questions and
are those that require a true/false answer. For example:
Rating scales are generally used to measure attitudes, opinions or feelings and
in a sense, these can be regarded as having a closed-item format because the
respondent has a limited range of options (ratings) to choose from.
form of closed question. This is a very simple kind of question that almost all
respondents should be able to deal with successfully. You could use this kind
of question when you want to obtain straightforward information, for example,
to find out what kind of hobbies people have. If you asked the respondents to
list their hobbies, you might get thousands of different ones. It may be easier
to give a list of hobbies and ask them to mark the ones that apply to them.
Here is an example:
sport
reading
cooking
watching television
stamp collecting
gardening
creative writing
art
other
You would not use this type of question format if you wanted to measure
attitudes or opinions, or the strength of people’s feelings about an issue. It is
merely a checklist, used in the same way that you would use a checklist to
make sure that you have packed everything you need for your holiday.
their answers to questions that follow (and this introduces bias). In addition,
the respondent who may be unsure of the best answer or misunderstands the
question, may select one of the alternatives randomly, without giving it much
thought.
2. Open questions
Open or free-response questions are not followed by any kind of choice and
the respondents provide their own answers that have to be recorded in full.
The main advantage of open questions is that respondents have the freedom to
express their ideas without the restrictions of set possible answers. Respondents
may have ideas and opinions that you have not thought of and these might be
lost if you ask a closed question. However, open-ended questions invariably
elicit some irrelevant and repetitious information that cannot be used and
wastes time in processing them. Answering open-ended questions also requires
a considerable degree of language proficiency and communication skills. For
this reason, they are not suitable for people with language difficulties or low
levels of literacy.
Although open-ended questions can give you a great deal of information, the
disadvantage is that they are difficult to analyse. If you want to compare the
responses of a large group of people, it may be better to use closed questions
for which all the respondents have the same choice of answers.
3. Rating scales
It is seldom useful to use single items or questions to measure complex or non-
factual topics such as opinions, beliefs, attitudes and values. These are complex
issues that have to do with states of mind rather than with behaviour or events
in the outside world and are therefore more difficult to measure. They are
usually multifaceted and have to be approached from different angles. Single
questions dealing with such topics are open to bias and unreliability due to the
way in which the questions are worded, the format of the questions and the
effect of the context regarding the way in which the questions are interpreted
and answered. Therefore, to measure non-factual topics, the tendency is to
use rating scales.
Rating scales here refer to multi-item scales, that is, a group of items dealing
with the same topic, each item requiring a rated response. For each item in
the rating scale, respondents have to indicate the extent to which they agree
or disagree with a statement by marking a point on a numerical scale. For
...........
122
R e s o u rce s
example, you could use a rating scale to investigate attitudes toward crime
in South Africa. Respondents would be required to rate their responses (by
ticking the appropriate box) to statements such as the following:
(Note that this can also be the case with other types of items.)
(1) Define the dimension being rated. This means that you have to decide
what it is that you want respondents to rate. Each item or statement to be
rated must refer to only one thing or dimension. For example, if you ask
respondents to “rate the shop assistant’s friendliness and efficiency” on a
scale of 1 (bad) to 5 (good), you are confusing two different dimensions:
friendliness and efficiency. The respondent might think that the shop
assistant was very efficient but not at all friendly. The respondent then
does not know which rating to use.
(2) Decide on the number of ratings for the scale. You may only need three
but there may be as many as 10. It depends on what is being rated. If
you only need respondents to indicate whether they agree, are neutral or
disagree, then you only need three ratings. If, however, you feel that there
may be a greater range of opinions, then you must provide more options
(a range of five or more). For example, look at the following statements
regarding people’s reading habits:
...........
123 PYC 2616/1
R E S O U R CE S
1 2 3 4 5
How often do you read books? Never One One a One a More than
in six month week one a
months week
These two ratings give you different information. The first one gives
you an indication of whether respondents read books or not, which is
fine if that is all you want to know. It does not tell you much about the
respondents’ reading habits. The second one gives you a much better
idea of their book reading habits. From this, you can see that the number
of ratings depends on what you want to find out.
(4) Define the different rating categories. You must specify criteria for each
rating so that they are mutually exclusive. This means that each rating
category should mean something different so that the respondents do
not have the problem of deciding which rating category their responses
fit into. For example, it might be confusing if your rating scale has the
following options:
Attitude scales are rating scales that consist of a group of items designed to
reflect different attitudes toward the topic in question. An attitude scale is
a technique for placing people on a continuum in relation to each other, in
relative and not in absolute terms. These scales are not designed to yield subtle
insights into individual situations and should not be used for clinical purposes.
Their main function is to classify people with respect to a certain attitude. This
allows the researcher to analyse the ways in which that attitude relates to other
variables in the survey. For example, you might find that different age or socio-
economic groups have different totals on a scale measuring attitudes to crime.
There are many different types of rating scales used to measure attitudes. We
will only discuss two here, namely Likert scales and the semantic differential.
...........
124
R e s o u rce s
1 2 3 4 5
Statement Disagree Agree
entirely entirely
Party XX has saved our
country from disaster.
Party XX is to blame for the
problems in the country.
I identify with party XX.
The respondent marks the point that best reflects his or her attitude. The
scores for each item or statement are then added up to obtain a total score for
the scale (and this is why it is referred to as a summated scale). When using
a summated scale, it is important to ensure that the scale is uni-dimensional,
that is, all the items measure the same dimension or topic. In this example, all
the items measure attitudes toward political party XX.
You will note that the statements in this example have been phrased in both
favourable and unfavourable terms. It is important to have both favourable
and unfavourable statements so that you do not influence the respondent in
any way. For example, if all your statements were phrased in a favourable
way, the respondents may tend to answer in a favourable light even though
they may think differently.
Likert scales usually have the option of five or seven ratings. The advantage is
that a number of ratings offer greater options than just a yes/no answer. There
are possible disadvantages to using an uneven number because people may
tend to choose the midpoint and then you do not know if the person is neutral,
lukewarm, lacks knowledge (about the content domain) or lacks an attitude
toward the issue in question.
...........
125 PYC 2616/1
R E S O U R CE S
Powerful _ _ _ _ _ _ _ Powerless
Hostile _ _ _ _ _ _ _ Friendly
Dynamic _ _ _ _ _ _ _ Static
Poor _ _ _ _ _ _ _ Rich
The respondent marks the rating that expresses his or her attitude. Using the
example of political party XX again, if the respondent thinks the party is very
powerful, he or she would place a mark on the first or second space closest
to “powerful”. If he or she thinks the party is powerful but not very much so,
then he or she is more likely to choose a space in the middle. If the respondent
thinks the party is really powerless, he or she would mark the space closest
to “powerless”, and so on. In this way, a picture can be obtained of people’s
attitudes toward the particular political party in terms of various descriptors.
When you compile a semantic differential scale, do not always put the positive
extreme on one side. The location of positive and negative poles should be
random to counteract any halo effect. The halo effect refers to the tendency
for respondents not to evaluate each item individually but for their responses
to be influenced by their general feeling of like or dislike. For example, having
decided on the direction of their attitude, people may tick down the one
side rather than reading each item if the items are arranged with favourable
responses on one side and unfavourable responses on the other.
It is also important that your two descriptors define the same construct. For
example, if you want to measure the construct of mood, your descriptors
might be happy/sad or stable/unstable. However, if your descriptors are happy/
reserved, it is not clear whether you are measuring the construct of mood
(happiness) or the ability to interact easily with others (reserved). In addition,
you should take care that the two descriptors really are opposites. For example,
sad/satisfied are not really opposites because you can feel satisfied about
something while still being sad.
The semantic differential is useful when you want to obtain an idea of people’s
endorsement of certain attributes. You could, for example, ask respondents
to rate the extent of their agreement that the organisation is powerful on a
scale of 1 to 5 but you would not get information about “powerful-ness” in
relation to the opposing attribute, that is, its “powerless-ness” unless you used
a semantic differential.
...........
126
R e s o u rce s
Layout of a questionnaire
(1) the name of the person or organisation conducting the study, to establish
credibility
(2) a general statement of the objectives of the questionnaire (people are
more likely to participate when they perceive that the study’s findings
will impact on them directly)
(3) assurance that their participation is valued and confidential (and in the
case of a survey questionnaire, that the answers are neither correct nor
incorrect)
(4) some estimate of the time required to complete the questionnaire
...........
127 PYC 2616/1
R E S O U R CE S
It is suggested that you take steps to ensure that no identifiable information will
be published without the person’s consent. The following may be displayed
prominently on the front of the questionnaire:
It is important to make sure that each question is directly relevant to the topic.
When you have drawn up your questionnaire, look at each question again
and ask yourself whether each one is necessary. If it seems too long, maybe
the same information can be obtained by combining some of the questions
or perhaps you need to be ruthless and just take out some of the questions, as
long as you are sure that you will still get the information you need.
However, shorter is not necessarily better. You need to have thorough coverage
of your topic in order to ensure “reliability” and “validity”. Sometimes it is
necessary to include two or more questions on the same topic to see if the
respondents are consistent in their answers. The aim is to strike a balance
between a concise questionnaire and one that is inclusive enough to ensure
validity.
...........
128
R e s o u rce s
(1) Try to avoid putting ideas into the respondents’ minds or suggesting
preferable attitudes. You might therefore start with open questions and
then introduce more structured questions at a later stage. For example, if
you are investigating people’s perceptions of the environmental impact
of nuclear power stations, you might begin with an open-ended question
such as, “What role does nuclear power play in our lives?” and then further
on in the questionnaire ask a more structured question, such as “What
kind of impact does a nuclear power station have on the environment?”
(Tick the appropriate answer.)
no impact at all
very little impact
a huge impact
If you had asked the question about nuclear power and the environment
first, this may have stimulated a particular perception based on an emo-
tional response (such as fear), for example that nuclear power stations
have a negative impact on the environment, without really thinking the
question through. The negative perception might then have influenced
the way the person responded to subsequent questions. By first asking
about the role of nuclear power, you are helping the person to think ra-
tionally about nuclear power and in this way, it is less likely that a certain
perception will be adopted toward the more emotive issues.
(2) Work from broad questions at the beginning to more specific questions
toward the end. This is called the funnel approach. In this way, you place
specific questions within a general context, which increases the likelihood
of obtaining meaningful responses. For example, in a questionnaire
about crime you might start with general questions about perceptions
of and attitudes toward crime and lead up to questions about personal
experiences or specific topics such as criminal justice or support systems
for victims of crime.
(5) When you are dealing with sensitive or very personal issues, these
should come toward the end of the questionnaire. You want to avoid
embarrassing or offending the respondents – if they are offended by
questions at the beginning of the questionnaire, they might then be
unwilling to complete the rest of the questionnaire. A well-designed
questionnaire is likely to make the respondent feel more comfortable about
answering sensitive questions that appear later on in the questionnaire.
(6) It is often useful to place one or more open-ended questions at the end
of the questionnaire to allow the respondents to express opinions or
feelings that are related to the purpose of the questionnaire but have not
been covered by the questions. You may think that you have covered all
relevant aspects of the topic but because individuals are unique, there is
always the possibility that you have not asked a question that pertains to
someone. By giving them the opportunity to add their own thoughts, you
will get better coverage of the topic. Furthermore, the respondents are
more likely to feel satisfied that answering the questions was worth the
effort. You could say something like “If you have additional information or
opinions that you would like to express, please do so here” and provide
space for the respondents to write in their comments.
...........
130
R e s o u rce s
and require a great deal of writing? The ideal is to vary the type of questions
so that the respondents do not get bored or irritated (which may affect the
validity of their responses).
6. Filter questions
If you have a group of related questions that apply to some people but not
to others, it is generally useful to start with a filter or screening question that
excludes some respondents from answering irrelevant questions. For example,
if you have a series of questions dealing with women’s magazines, it is likely
that some of your respondents do not buy women’s magazines and therefore
several questions will not apply to them and consequently there is no point
in answering further questions about them. You could begin by asking for
factual information by, for example, asking them to indicate on a checklist
what magazines they buy. A typical filter question would then be “Have you
bought a women’s magazine in the last three months?”. If the answer is no,
the respondent is asked to skip the next few questions that deal with women’s
magazines.
...........
131 PYC 2616/1
R E S O U R CE S
1. Purpose of a manual
Someone else might be interested in using your test or questionnaire. We
will refer to this person as the user of the questionnaire. You need to write a
manual for your questionnaire so that this person can see if the questionnaire
is relevant for the specific purpose he or she wants to use it for. You should
clearly state the aim of the questionnaire and describe the target population.
Relevant information on the use of the questionnaire in a multicultural and
multilingual context should be given and the cultural and language group/s for
which the questionnaire is suitable should be specified. A brief description of
the design of the questionnaire helps the user to understand how this design
ensures that the questionnaire can be used for the intended purpose.
Once the user has decided that the questionnaire is relevant, he or she would
like to know how well this questionnaire would work for the intended purpose.
To enable users to evaluate the properties of the questionnaire, you should
report the process of analysing and selecting the items as well as the reliability
and validity of the questionnaire. If the questionnaire is to be used for more
than one cultural and/or language group, equivalence for these groups in terms
of reliability and validity should also be discussed. Based on this information,
the users can decide for themselves if the properties of the questionnaire are
acceptable or if there is another questionnaire that might provide better (i.e.,
more reliable and valid) results.
2. Structure of a manual
The following outline can be used to structure the manual:
Aim
Target population
Design of the questionnaire
...........
132
R e s o u rce s
The aim of the questionnaire determines for whom it will be used. You should
describe those characteristics of the target population that are relevant to the aim
of the questionnaire. Remember that the user should have enough information
to be able to decide if the questionnaire is relevant for his or her purpose. The
questionnaire on food expenditure would probably be targeted at a national
sample that includes different cultures and languages. It would be more suitable
for adults. In the case of the attitude questionnaire it would be important to state
for which country this questionnaire has been developed (different countries
have different systems of law). Another important characteristic would be age
as the subject matter of this questionnaire (the death penalty) makes it more
suitable for adults.
Once the aim has been stated and the target population has been defined, a
brief description of the design of the questionnaire should be provided. The
purpose of the questionnaire determines the content domain and it should
be described how the items of the questionnaire cover this domain. The type
of items should also be indicated, for example multiple-choice items on food
expenditure where the respondent selects the relevant response from a number
of alternatives. In the case of the attitude questionnaire, the respondent might
have to rate how much he or she agrees with a statement about the death
penalty.
...........
133 PYC 2616/1
R E S O U R CE S
they are representative in terms of those characteristics that define the target
population. It should also be mentioned when and under which circumstances
the questionnaire was administered to them.
You should briefly describe each technique used for item analysis and you
should indicate which criteria were used to justify the inclusion or exclusion
of items in the item selection process. For example, to determine if the items
in a questionnaire all measure the same thing, one will look at the relation
between each item and the total of all the items (see also “item analysis”). A
strong relation indicates that the specific item measures the same thing as the
other items in the questionnaire. If this is important, one may decide to include
only those items with a moderate to strong relation with the total.
It is furthermore important for the user to know how reliable or consistent the
questionnaire is for the people for whom it is to be used. You should give a
brief description of the method used to determine reliability and justify why
this type of reliability was used (as opposed to the other types of reliability).
The reliability coefficient is then evaluated in terms of what can be regarded
as an acceptable level of reliability given the purpose of the questionnaire.
The reliability should be acceptable for all the cultural and/or language groups
for whom the questionnaire will be used.
A questionnaire is valid for a specific purpose and you need to identify the
category of validity (be it content validity, criterion-related validity or construct
validity) that is relevant for your questionnaire. You must explain how the
evidence was gathered to determine the validity, and based on this evidence,
you need to discuss the extent to which the questionnaire measures what
it claims to measure. In other words, you must evaluate the validity of the
questionnaire. The questionnaire needs to be equivalent for all the cultural
and/or language groups for whom the questionnaire will be used. Prediction
bias refers to a lack of equivalence in terms of criterion-related validity and
construct bias refers to a lack of equivalence in terms of construct validity.
You should provide instructions for scoring so that a numerical value can be
given to the answers on those items where scoring is relevant. In the case of
...........
134
R e s o u rce s
The guidelines for interpretation of the results should be based on the aim
of the questionnaire. If a questionnaire (or part of a questionnaire) measures
how well someone has mastered a course on managing household finances,
then the total score indicates the individual’s knowledge of the course content.
Someone with a high score obviously knows more than someone with a low
score. If the statements in a rating scale all indicate problems with the death
penalty as means of punishment, then a high total score could be interpreted
as that the individual does not regard the death penalty as an acceptable means
of punishment. Because respondents can agree (or disagree) to a greater or
lesser extent with the statements on a rating scale, one can distinguish between
individuals in terms of their attitude. One could also interpret an individual’s
total score by comparing it to the average total score of a group. The average
total score of a group is calculated by adding up the total scores for all the
individuals in the group and dividing the sum by the number of individuals.
...........
135 PYC 2616/1
R E S O U R CE S
Reliability
Not only are these conditions irrelevant to the purpose of the questionnaire,
they might also differ from one administration to the next. Their effect on the
results of the questionnaire is therefore unpredictable and inconsistent. These
irrelevant conditions are called unsystematic sources of variation. To vary
means to change. The reliability of a questionnaire refers to how consistently it
measures regardless of the occasion on which it is administered or the sample
of items included. It therefore refers to the consistency of results over different
administrations involving different occasions, questionnaire forms, etcetera.
...........
136
R e s o u rce s
...........
137 PYC 2616/1
R E S O U R CE S
If several days have elapsed between the two administrations (of the two forms),
alternate-forms reliability is not only a measure of the consistency of results
over different forms but it can also be used to determine the consistency of
results over different occasions or stability over time. This offers some solution
to possible memory effects experienced with test-retest reliability.
You now have an estimate of the reliability of each half of the questionnaire
(it is the same for each half) and not of the whole questionnaire. This is
probably an underestimation of the reliability of the full questionnaire as a
shorter questionnaire is generally less reliable than a longer questionnaire. A
correction is made to obtain the reliability of the whole questionnaire. This is
called the split-half reliability of the questionnaire and it measures the degree
of equivalence between the two halves of the questionnaire, that is, the extent
to which they measure the same attribute. As such, it reflects the consistency
of the group’s performance on the items within the questionnaire and indicates
the degree of relatedness of the items. This type of reliability is therefore
regarded as a measure of the internal consistency of the questionnaire and
the closer the split-half reliability is to 1, the higher the internal consistency
of the questionnaire.
...........
138
R e s o u rce s
3. Evaluating reliability
It is important to note that it is not correct to refer only to the reliability of a
questionnaire. The questionnaire you are dealing with should be specified
together with the representative group for whom the estimate of reliability
has been determined. The nature and purpose of the questionnaire determine
which type of reliability is appropriate and the acceptable level of reliability
also depends on the use of the questionnaire. If you are working with a
psychological test of which the results could have important consequences
for an individual, such as an intelligence test, the reliability coefficient should
be above 0.90. However, a questionnaire with a reliability coefficient of 0.70
(and even somewhat lower) can be useful if the results are used in combination
with other information about the individual or group.
...........
139 PYC 2616/1
R E S O U R CE S
Specification document
Before you start compiling a questionnaire, you should have a rough idea of
the line of enquiry you wish to follow. Do you need short, factual answers
or do you want to conduct analytical research on a set of attitudes? How
many people will be involved? Are they children or adults? If adults, are they
businesspersons or the unemployed? All these questions influence the kind
of questions you will ask, the level of language you use, how complex the
questions are and so on. In this way, the purpose of the investigation, the kind
of information you want and the characteristics of the respondents (that is, the
target population) all influence the way the questionnaire is compiled, that is,
they influence the questionnaire specifications. The detailed specification of
measurement aims should be clearly related to the purpose of the research.
...........
140
R e s o u rce s
Suitability of a questionnaire
There are many different instruments and techniques for gathering information
from people. You can use questionnaires, standardised ability and achievement
tests, projective tests, interviews, observation, narrative discourse and a range
of different methods. The measuring instrument and approach you use depend
on the topic you have chosen and the purpose of your investigation.
Exercise
In the following exercise, you are given a list of topics for investigation (left
column). In the middle column, indicate whether you think a questionnaire
would be a suitable measuring instrument and in the last column, give a reason/
reasons for your answer. Sometimes you just have to rely on your common sense
to decide whether you can get the information you want with a questionnaire
or not but you should always be able to give a reason for your decision.
...........
141 PYC 2616/1
R E S O U R CE S
Comments on exercise
(1) Yes.
You want to find out facts. Your questionnaire could include questions
about the parties’ respondents support, or ratings of various aspects of
the party and their policies, etcetera.
(2) Yes.
Again, you are looking for factual information. You might have a list of
the different beers and their qualities and ask people to indicate what
they like and why.
(3) No.
Here you are looking at a practical ability and in order to assess typing
skills, you would have to administer a practical test of typing ability. You
could use a questionnaire if you wanted facts about typing qualifications,
work experience, etcetera.
(4) Yes.
Questionnaires are frequently used to gather information about people’s
opinions on various issues.
(5) Yes.
This is a slightly tricky one. You can use a questionnaire to find out
about parenting practices if you ask, for example, about attitudes toward
punishment, about routines in the home, food provided and stimulation
available. On the other hand, if you wanted to examine the effectiveness
of certain parenting practices, you would use observation rather than a
questionnaire.
...........
142
R e s o u rce s
Validity
1. Content validity
The purpose of the questionnaire refers to what it intends to measure and for
whom it is to be used. This in turn determines the content domain the researcher
needs to cover. If the purpose of the questionnaire is to measure achievement
in a specific course (that is, how well a student has mastered all the objectives
of the course), the content domain is the course material and related tasks
the student should be able to perform. The purpose of a questionnaire could
also be to measure behaviour in a particular context and the content domain
would then be different possible behaviours in that context. For example, if
you want to measure behaviour in a crowd situation, the content domain you
need to cover includes different behaviours such as losing one’s inhibitions
or on the other hand becoming withdrawn.
...........
143 PYC 2616/1
R E S O U R CE S
Content validity should not be confused with what is called face validity. Face
validity refers to the degree to which items in a questionnaire appear to be
relevant to what is being measured. Face validity is based on the subjective
evaluation by people who are not necessarily experts either in the particular
area or in psychometrics. If the respondents do not regard the items as relevant
(the questionnaire does not have sufficient face validity), they might be less
motivated and even unwilling to cooperate.
2. Criterion-related validity
Tests and questionnaires are often used to estimate an individual’s position or
performance on some outcome measure. This outcome measure is referred
to as the criterion. The criterion-related validity of a questionnaire is the
extent to which the scores on the questionnaire are effective in estimating an
individual’s position or performance on the relevant criterion. Two different
approaches to gathering evidence of criterion-related validity are concurrent
validity and predictive validity.
...........
144
R e s o u rce s
3. Construct validity
A construct is an unobservable quality which forms part of a theory designed
to explain observable behaviour. For example, anxiety (a construct) is not
observable but it forms part of a theory that explains observable behaviours
such as nail biting, sleep disturbances and eating disorders. Constructs influence
behaviour and because you cannot directly observe a construct, you infer it
from the behaviour associated with that construct. If you see someone showing
the above behaviours, you may infer that this person has some anxiety.
Based on the relevant theory, you would expect groups who are supposed
to differ in terms of a construct to also obtain significantly different scores
on a questionnaire measuring this construct. If your questionnaire measures
the degree of warmth someone shows towards others (interest in people,
outgoingness, etc), then you could administer the questionnaire to a group of
social workers and to a group of research scientists as part of the validation
process. You would expect the social workers to obtain a significantly higher
average score on the questionnaire indicating that they possess a higher degree
of the construct. Such a result would provide support for the construct validity
of the questionnaire and for the theory underlying the construct.
...........
146
R e s o u rce s
Let us assume that you want to find out whether Unisa students feel lonelier
than residential students do and that you find an existing questionnaire that
appears to contain a number of highly relevant items. However, a few are
unsuitable as they relate to loneliness in the work situation. You may choose
to eliminate these, or you may decide to modify them. For example, you may
change the question “How lonely are you at work these days?” to “How lonely
are you at university?” (Be careful: If you eliminate or modify items, you should
not report the reliability and validity estimates of the questionnaire published
by the original author as these properties might be different for a questionnaire
containing the “changed” items.)
The guidelines presented below advise you on what to take into account
when constructing or adapting items. It is important to note that they are only
guidelines – each questionnaire is unique and presents its own demands and
problems.
• Do read each item and ask yourself if the item relates to your topic. Does it
tap the construct you want to measure? For example, if you want to determine
whether pupils perceive their science teacher as a “good” teacher don’t ask,
“Do you like your science teacher?” Likeability is not the same as capability.
• Don’t be tempted to ask questions that are interesting but not vital to your
research. Be ruthless in disregarding such items.
may misinterpret items written in such language, or simply choose any answer
in order to avoid appearing stupid. This will invalidate your results.
• Don’t use academic or technical terminology, jargon, words that are seldom
used in everyday speech, very long sentences, or complicated syntax (see
example below).
• Do phrase your items in such a way that the language level matches that of
your respondents. All the words in your items should be familiar to them.
For example, the following item may be easily understood by teachers but
not by 12-year olds: “The notion that pedagogues are habitually unfair to
scholars is preposterous”. Rather state: “Teachers are often unfair to pupils”.
• If you are not sure whether items would be easily understood by your
target population then do present them to a small group of respondents
from your target group and ask these people to “think aloud” while they
respond to them.
3.4 Clarity
All instructions, questions and statements in a questionnaire must be so clear
that all the respondents interpret them in the same way and understand
...........
149 PYC 2616/1
R E S O U R CE S
Writing clear instructions and items is not always easy. It often requires
painstaking effort, pre-tests and revisions to ensure clarity. If anything in your
questionnaire is not understood and/or misinterpreted, your results will be
useless.
• Don’t ask questions with two inherent issues. If there is more than one
issue inherent in an item the respondents will not know which to answer.
Consider these examples:
I am fully occupied and I don’t feel lonely.
I don’t feel lonely because I am fully occupied.
Do you find the psychology course informative or interesting?
For example, imagine you (a) are fully occupied (you have a full-time job
and are studying through Unisa), and (b) feel extremely lonely because you
do not have any time to socialise and/or do not have any close friends or
relatives. How would you answer the first two items above? Rather break
such questions or statements into two separate items.
• Do scrutinise any items that contain the conjunctions “and” or “or” to see
if they contain more than one possible issue.
...........
150
R e s o u rce s
• Do use active rather than passive statements. Passive statements are more
difficult to understand, and therefore more difficult to respond to, than
active statements. Here is an example of a passive statement: It is believed
by students that they will be given extension by lecturers.
The following is simpler:
Students believe lecturers will give them extension.
• So do write items that are specific, simple, clear, and to the point.
3.5 Configuration
“Closed” items offer a given choice of responses. When you construct a set
of such items, be sure that the given responses are appropriate for each item.
In other words, your item should be configured correctly.
(a) Have you met any other students who are studying psychology?
Now suppose students are asked to respond “strongly agree”, “agree”, “uncertain”,
“disagree”, or “strongly disagree” to a set of items. Which of these responses
could they give to the following statement?
...........
151 PYC 2616/1
R E S O U R CE S
Not all the given responses above are appropriate to this statement. “Uncertain”,
“strongly agree” and “strongly disagree” are unnecessary. All the given responses
would be more suitable to a statement such as “I believe one should phone
a lecturer whenever one doesn’t understand something in the textbook”.
Indeed “always”, etcetera would be a more suitable set of given responses to
question (b).
• Do limit the time frame to the immediate past (such as the last week – at
the most the last six months) when asking about what happened in the past.
If the respondents have to spend a lot of time and effort to remember what
they did or felt in the distant past they are likely to guess or give up. And
this will affect the reliability of the information they provide. For example,
it may be difficult for a student to answer the following question: “How
many hours did you spend on doing assignments last year?”
• When asking questions relating to factual information, do make sure your
respondents have the information. For example, if you ask students “Do
you agree or disagree with Unisa’s policy regarding entrance requirements?”
make sure your respondents are aware of the policy. In such a case it is
better to first ask whether they are aware of the policy.
• Don’t give examples unless it is really necessary, as these too may influence
certain respondents. For example:
Do you use any word processing packages, such as ZZ?
Rather list all possible programs or use no examples at all.
...........
152
R e s o u rce s
Faking can be reduced if the respondents are made aware that they are not
being individually assessed and that their responses will remain anonymous.
...........
153 PYC 2616/1