LITTLE

QUICK FIX:
WRITE A
QUESTIONNAIRE

#LittleQuickFix
Sara Miller McCune founded SAGE Publishing in 1965 to support
the dissemination of usable knowledge and educate a global
community. SAGE publishes more than 1000 journals and over
800 new books each year, spanning a wide range of subject areas.
Our growing selection of library products includes archives, data,
case studies and video. SAGE remains majority owned by our
founder and after her lifetime will become owned by a charitable
trust that secures the company’s continued independence.

Los Angeles | London | New Delhi | Singapore | Washington DC | Melbourne

 LITTLE
QUICK FIX:
WRITE A
QUESTIONNAIRE

Helen
Kara
SAGE Publications Ltd © Helen Kara 2019
1 Oliver’s Yard
55 City Road First published 2019
London EC1Y 1SP

SAGE Publications Inc.

2455 Teller Road
Thousand Oaks, California 91320

SAGE Publications India Pvt Ltd

B 1/I 1 Mohan Cooperative Industrial Area Apart from any fair dealing for the purposes of
Mathura Road research or private study, or criticism or review, as
New Delhi 110 044 permitted under the Copyright, Designs and Patents
Act, 1988, this publication may be reproduced, stored
SAGE Publications Asia-Pacific Pte Ltd or transmitted in any form, or by any means, only with
3 Church Street the prior permission in writing of the publishers, or in
#10-04 Samsung Hub the case of reprographic reproduction, in accordance
Singapore 049483 with the terms of licences issued by the Copyright
Licensing Agency. Enquiries concerning reproduction
outside those terms should be sent to the publishers.

Editor: Aly Owen

Production editor: Ian Antcliff
Marketing manager: Ben Griffin-Sherwood Library of Congress Control Number: 2018956503
Design: Lisa Harper-Wells
Typeset by: C&M Digitals (P) Ltd, Chennai, India British Library Cataloguing in Publication data
Printed in the UK

A catalogue record for this book is available from

the British Library

ISBN 978-1-5264-6775-1 (pbk)

At SAGE we take sustainability seriously. Most of our products are printed in the UK using responsibly
sourced papers and boards. When we print overseas we ensure sustainable papers are used as measured
by the PREPS grading system. We undertake an annual audit to monitor our sustainability.
Contents
 Everything in this book!������������������������������ 4

Section 1 How do I know I should use

questionnaires in my research project?�������� 9

Section 2 What are the practicalities of using

questionnaires?������������������������������������������ 23

Section 3 What kinds of questions should I use?������� 37

Section 4 How do I find people to fill in my

questionnaires?������������������������������������������ 63

Section 5 Why do I need to test my

questionnaire?�������������������������������������������� 79

Section 6 How can I improve my response rate?������� 93

Section 7 What do I do when I have all my

completed questionnaires?���������������������� 107

Glossary��������������������������������������������������� 118

3
 2
MIN

summar
y

Everything in
this book!

Section 1 First you need to know whether questionnaires are right

for your project. Are you researching a topic that is not too personal
or sensitive? And do you want specific, measurable information? If so,
questionnaires are likely to work for you.

Section 2 The practicalities of using questionnaires depend on how

you decide to administer them – on paper, online, by text message, or
on a device such as a tablet.

Section 3 You can use any or all of a wide variety of question types,
such as demographic questions, yes/no questions, dichotomous
questions, multiple-choice questions, scaled questions, and open
questions.

4
Section 4 You can find people via postal or email address lists, in
person, or online.

Section 5 Every questionnaire should be tested to find out how well

each individual question works and how well the questionnaire works as
a whole.

Section 6 Some options for increasing your response rate include

follow-ups (for posted or emailed questionnaires), give-aways (for
questionnaires administered by post or in person), incentives, and
visually interesting questionnaires.

Section 7 When you have all your completed questionnaires you need
to prepare your data for analysis.

5
6
Section

Decide whether
questionnaires
are right for
your project

7
How do I know
I should use
questionnaires
in my research
project?
 10
SEC

summary

When you want specific,

measurable information
about a topic that is not
too personal or sensitive.
 60
SEC

summary

Questionnaires are not as simple as they look!

A questionnaire is just a set of written questions you can use to find out
about stuff, right?

Wrong!

A questionnaire is a carefully constructed research instrument made

up of tested questions designed to gather specific information. The
information will be about a topic that is not particularly personal or
sensitive. Most if not all of the information gathered will be measurable,
and will enable you to compare responses. If this is the kind of
information you need, then you should consider using questionnaires.

The other thing you should consider is whether your intended

respondents are likely to be able to read. It is not sensible to try to use
questionnaires with, for example, young children or people with visual
impairment.

11
 WHAT IS A
QUESTIONNAIRE
ANYWAY?

A questionnaire is a method of collecting data from individuals using

writing. It contains different types of questions, and may be produced
on paper, a device such as a tablet, by text message, or online.
The people who answer questionnaires are known as ‘respondents’
(because they respond to the questions).

A research questionnaire is different from an online poll or a

magazine quiz. Questionnaire questions, and their answers, are
carefully designed and tested to help you find out what you need to
know to answer your research question.

Also, a questionnaire is different from a survey, which is the systematic

collection and analysis of data. Questionnaires may be used in survey
research, though it is not essential; another method such as interviews
may be used.

12
 WHEN WOULD
I USE ONE?

A questionnaire is useful if:

• You are researching a topic that is not personal or sensitive

• You want specific and measurable information
• Your respondents are likely to be able to read and understand the
questions
• You want to compare responses from different categories of people

13
WHAT IS A SUITA
B
FOR A QUESTIONLE TOPIC
NAIRE?

Questionnaires are most useful for factual, everyday topics. If you

want to research people’s views of a service or event, questionnaires
are a good method to use. If you want to research something
personal or contentious, you should consider a different method such
as interviews. This is because people are more likely to give fuller
responses to difficult questions if they’re talking to another person.
 WHAT INFORMAT
IO
CAN I GET FROM N
QUESTIONNAIRE
S?

Questionnaires generate specific and measurable information within

a clear structure. You can ask questions about the demographic
characteristics of your respondents, such as their age, gender, or
ethnicity. Then you will ask questions about your area of interest. Most
are likely to be closed questions, i.e., having pre-defined answers.

Closed questions are useful because they generate specific and

measurable information. You will need to devise a range of answers to
a question for your respondents to choose from. Questionnaires may
also include open questions. However, you should keep these to a
minimum, because they are harder for respondents to answer and for
you to analyse. Ultimately, questionnaires will help you to tell the story
of your research question and its answer.
 ER
WHAT TO CONSID L
ABOUT POTENTIA
RESPONDENTS

Your respondents need to be able to read. This rules out, for example,
very young children and people with serious visual impairment. Then
they need to be able to read the language in which you’re writing the
questionnaire. And they need to be able to understand what you have
written. This may rule out young children, some people with cognitive
impairment such as from dementia or brain injury, and some people with
learning disabilities.

If you want to distribute your questionnaire online, your respondents

need to have internet access and the skills to respond to an online
questionnaire. Similarly, if you want to use text messaging, your
respondents need to have mobile phones and the ability to use them.

16
 ARE
HOW CAN I COMP
RESPONSES?

The demographic information you collect can be used to compare

and contrast the answers to other questions. For example, do men,
women, trans or non-binary people give different answers? Also, you
can compare different answers about your area of interest. For example,
if you’re studying public transport: do people feel differently about bus
and train travel? Comparing responses isn’t essential, but if you want to
do this, a questionnaire can be helpful.

17
 QUESTIONNAIRE...
Or Not?

Are you researching a factual everyday topic that

isn’t personal or sensitive? YES / NO

Do you want to collect information that is specific

and measurable? YES / NO

Are your potential respondents likely to be able to read? YES / NO

s’ to each of these,
If you can answer ‘ye
ld be a good method
then questionnaires wou ta
for collecting your da

18
#LittleQuickFix

19
 Decide how
to administer
questionnaires

20
Section
21
 What are the
practicalities
of using
questionnaires?
 10
SEC

summary

Practicalities include how you will

administer questionnaires, how
questionnaires may be completed
and returned, and the nature of the
questionnaire itself.
 60
SEC

summary

What do I need to think about?

We have a range of options for how to administer a questionnaire:

• On paper
• Online
• By text message
• On a device such as a tablet

Questionnaires can be completed by respondents in your presence,

over the phone or internet, or at a distance in their own time. They may
be long or short, simple or complex.

It is important to ensure that questionnaires are anonymous to protect

your respondents. If you need to collect personal information for any
reason, it should be kept separate from the completed questionnaires.
 QUESTIONNAIRE
S
PAPER OR ONLINON
E

Paper is useful when you’re working with respondents who don’t

use the internet, such as some older people. It is also useful in areas
where internet connectivity is poor or non-existent, such as some rural,
coastal, or desert regions. But it is laborious to prepare the data for
analysis by entering it into a computer.

Most questionnaires are created and administered online. Software

such as SurveyMonkey or QuestionPro offers great flexibility and
convenience to researchers and respondents. It allows researchers to
create questionnaires with a range of pathways, while respondents only
see the questions they are intended to answer. Also, data arrives ready
for analysis. However, online questionnaires are only useful when your
respondents have internet access and the necessary online skills.

26
 QUESTIONNAIRE
SB
TEXT MESSAGE Y
OR ON A DEVICE
These methods are less common, but useful for some people.
Software such as SmartSurvey and SMS-Track is available online
for questionnaires to be administered by text message. This is quick
and easy for many respondents. But of course it only works with
respondents who have smartphones, and whose numbers you have or
can obtain. Also, you have to keep your questionnaires very short or
split them into sections.

Another option is to create a questionnaire on a device such as a tablet

that you can offer to people to complete in your presence. This is a
technological replacement for the paper questionnaire on a clipboard,
and can be useful if you want to find respondents in public places such
as train stations or sporting events.
 LD MY
HOW LONG SHOUE BE?
QUESTIONNAIR

Questionnaires may be:

• Long
• Short
• Simple
• Complex
• Inbetween

Longer questionnaires generate more data, but are harder for people to
complete. Generally, it is worth keeping your questionnaire as short
and simple as possible. If you’re working on paper, aim for two sides
of A4 at most, and a good-sized font; 12-point minimum, 14-point to
cater for the needs of any visually impaired people. If you’re using text
messages, you’ll need to keep your questionnaire short. Online or on a
device it can, in theory, be as long as you like. However, to avoid over-
burdening respondents, it’s probably best to think in terms of 10–12
questions at most for each person to answer.

28
 HOW CAN
S
QUESTIONNAIRE
BE COMPLETED
AND RETURNED?

Questionnaires may be completed:

• With the researcher in person

• With the researcher on the phone
• With the researcher online
• At a distance in the respondent’s own time

Questionnaires completed at a distance may

be returned:

• By post
• By email
• Online
• By text message

29
 THE IMPORTANCE
OF ANONYMITY

It is important that research participants can be anonymous. This means

you shouldn’t ask people to include their names, addresses, phone
numbers, or any other identifying information in their responses.
The exception is if you are offering an incentive (see below for more on
this). In such cases you could ask for contact information separately
from the questionnaire. For example, you could use a separate piece
of paper for people to write on if you’re administering questionnaires
in person. For online questionnaires you could give an email address
for them to send the information to. Alternatively you can collect the
information with the questionnaire in such a way that you can separate
it immediately.

30
31
 FIGURE OUT YOUR
PRACTICALITIES!

ers
Write in your answ
e are
In some places ther
clues to guide you

I will administer my questionnaires

…………………………………………………….
(on paper/online/by text message/on a device)

Because ……………………………………………

……………………………………………………

……………………………………………………

……………………………………………………

32
 CHECKPOINT

Respondents will complete the questionnaires

…………………………………………………….
(with me in person, on the phone, or online, or by
themselves at a distance)

Questionnaires will be returned to me

…………………………………………………….
(by post, by email, online, in person)

I will ensure participants’ anonymity by

…………………………………………………….

33
 Use any or all
of a variety of
question types

Section

34
35
36
 What kinds
of questions
should I use?
 10
SEC

summary

You can choose from

demographic questions, yes/
no questions, dichotomous
questions, multiple-choice
questions, scaled questions,
and open questions.

38
 60
SEC

summary
Questions, questions, questions!

There are many kinds of questions available for use in questionnaires.

We have included details of the six most common kinds here. These
are:

• Questions to collect information about the people in your sample,

called demographic questions
• Questions to classify those people into groups: yes/no questions
• Questions to find out about respondents’ knowledge or experience:
- Dichotomous questions
- Multiple-choice questions
- Scaled questions
- Open questions

This section will take you through each type of question step by step.
It will:

• Explain each question type

• Give you one or more examples
• Offer an interactive exercise or two
• Discuss the relevant features of the question type, and
• Give you the opportunity to design questions of that type for
your own research.
39
 QUESTION TYPE 1:
DEMOGRAPHIC
QUESTIONS

Classifying characteristics

Questionnaires often start, and sometimes finish, with demographic

questions. These questions aim to classify respondents on the basis
of certain characteristics that are of interest to the researcher. Age and
gender are commonly used, though not compulsory. To make life easier
for respondents, it is usual for researchers to provide pre-determined
answers.

40
What do you think of these demographic questions? Write down your
thoughts below each question, then turn over for the answers.

Age: 0–5 5–20 20–50 50–60 60+

…………………………………………………….

…………………………………………………….

……………………………………………………..

…………………………………………………….

Gender: M/F
……………………………………………………..

…………………………………………………….

……………………………………………………..

…………………………………………………….

41
 Age questions

The age question is problematic for two reasons. First, the intervals
are uneven: five years between 0-5, 15 between 5-20, 30 between
20-50 and so on. Second, the age bands should start and finish with
different numbers. Where would someone aged 20 place themselves
in these bands? They have two options. That means the data would
be unreliable, as the researcher wouldn’t know how many 20-year-old
respondents had chosen 5-20 and how many had chosen 20-50.

Can you write a better version of the age question for your own
research?

……………………………………………………..

…………………………………………………….

……………………………………………………..

42
Gender questions

The gender question is old-fashioned as we now know there are more

than two genders. Trans people are often separated out in gender
questions, and non-binary people may be included. There are a number
of other categories such as genderqueer and genderfluid. In 2014,
Facebook offered its users over 50 gender categories, which you can
find online.

How would you formulate a gender question for your research?

……………………………………………………..

…………………………………………………….

……………………………………………………..

43
 Other common demographic questions

Demographic questions need not be limited to age and gender. Other

common options include ethnicity and marital status. Depending on
your area of study, you might ask a demographic question about hair
colour or voting frequency. Can you think of one, two or three other
demographic questions that would be useful for your own research?

1 .……………………………………………………..

2 ……………………………………………………..

3 ……………………………………………………..

How would you structure the pre-determined answers?

1 .……………………………………………………..

2 ……………………………………………………..

3 ……………………………………………………..

44
QUESTION TYPE 2:
YES/NO QUESTIONS

Classifying interests

Like demographic questions, yes/no questions can help to classify

people into groups. However, unlike most demographic questions,
those groups can be related to your areas of interest. These questions
are very useful for researchers, though they can be frustrating for
respondents. Take this example from a study of healthy eating habits:

Do you like tomatoes? Yes / No

What do you think of this question? Write down your thoughts then turn
over for the discussion.

……………………………………………………..

…………………………………………………….

……………………………………………………..

45
 Intermediate answers

Let’s say a respondent likes cooked tomatoes but can’t bear raw
tomatoes. How would they answer? For this reason, some researchers
argue that we should always give an intermediate answer such as
‘sometimes’ or ‘maybe’.

Intermediate answers can help to increase response rates, but

they make analysis more difficult because you have three categories
to deal with. Also, if someone answers that they like tomatoes
‘sometimes’, you don’t know anything about the conditions under which
they like or dislike tomatoes.

Can you think of one, two or three yes/no questions that would be
useful for your own research? Add an intermediate answer if you wish.

1 .……………………………………………………..

2 ……………………………………………………..

3 ……………………………………………………..

46
QUESTION TYPE 3:
DICHOTOMOUS
QUESTIONS

Classifying between choices

Yes/no questions are an example of dichotomous questions. Other such

questions can be used to ask about feelings and attitudes. The answers
you offer could be happy/unhappy, pleased/displeased, or any such
pairing. Here’s an example from research into the experiences of team
sport fans:

How do you feel about your team’s latest game? Satisfied/unsatisfied

What do you think of this question? Write down your thoughts then turn
over for the discussion.

……………………………………………………..

…………………………………………………….

……………………………………………………..

47
 Pros and cons of dichotomous questions

Again, dichotomous questions are useful for researchers because

they make analysis easier, but they can be frustrating for
respondents. Imagine someone thought their team performed better
than the other team, but lost the game because they believed the
referee or umpire was incompetent. Would that person be satisfied or
unsatisfied?

As with yes/no questions, you can offer an intermediate answer such as

‘neither’. But again this makes analysis more complicated.

Can you think of one, two or three dichotomous questions that would be
useful for your own research? Add an intermediate answer if you wish.

1 .……………………………………………………..

2 ……………………………………………………..

3 ……………………………………………………..

48
QUESTION TYPE 4:
MULTIPLE-CHOICE
QUESTIONS

Classifying preferences

These are questions that offer respondents several options to

choose from. They are not like multiple-choice questions in exams
where some answers are false. They may be constructed so that the
respondent should only choose one answer, or so that the respondent
can choose more than one answer. Here are two examples:

What is your favourite ice-cream Which ice-cream flavours do you

flavour? Tick one only. like? Tick as many as apply.

� Vanilla � Vanilla
� Chocolate � Chocolate
� Strawberry � Strawberry
� Salted caramel � Salted caramel
� Mint choc chip � Mint choc chip

However, there are at least two potential problems with these questions.
Can you identify them? Write your answers below, then turn the page.

…………………………………………………….

……………………………………………………..
49
 Challenges of multiple-choice questions

What if the respondent’s favourite ice-cream flavour is tutti frutti, or

cookies and cream, or jalapeño? And what if they don’t like ice-cream
at all?

It can be very difficult for a researcher to pre-determine all the

possible answers that a respondent could give. One way to help
resolve this is to add an option for ‘none of the above’. Another is to
add a box marked ‘other’ and let respondents tell you in their own
words. This will yield more accurate data, and be less frustrating for
your respondents, but of course your data will be more complicated to
analyse.

Can you think of one or two multiple-choice questions that would be

useful for your own research? Add a box marked ‘other’ if you wish.

1 .……………………………………………………..

2 ……………………………………………………..

50
QUESTION TYPE 5:
SCALED QUESTIONS

Classifying levels

These are questions with scales for answers and the respondent is
asked to select a point on the scale. Examples of scaled questions are:

Overall, how satisfied are you with the event you attended?

� Very satisfied
� Quite satisfied
� Neither satisfied nor dissatisfied
� Quite unsatisfied
� Very unsatisfied

How capable do you feel when writing essays?

Perfectly capable 10 9 8 7 6 5 4 3 2 1 Totally incapable

You may spot some key differences between these two questions. Try
to identify them, then read the discussion that follows.

……………………………………………………..

51
 Scaled question structures

The first question has a scale made of words, the second has a scale
made of numbers. Whichever you use, the distance between each point
on the scale should be the same or as close as possible.

Also, the first question has an odd-numbered scale with five options,
while the second has an even-numbered scale with 10. Why do you
think this matters?

.……………………………………………………...........

……………………………………………………............

choose one side or the other.

answer, while an even-numbered scale forces them to
An odd-numbered scale offers respondents an intermediate
Answer

52
Choosing a scale type

Whenever you are using a scaled question, you have to decide which
kind of scale to use. An even-numbered scale with a lot of response
options, as in the second example question, reduces the impact of
being forced to make a choice. The respondent can choose to be ‘only
just’ on one side or the other. You can have as many points as you like,
though it is easiest for respondents if you don’t have too many. Having
said that, if you find at a later stage that there are too many, you can
collapse them into smaller categories. If you don’t have enough, you
can’t expand them.

Digital questionnaires have a great advantage over paper questions

when it comes to scaled questions: respondents can’t choose a point
between two numbers. With paper questionnaires they can – and do –
which makes analysis more difficult.

Can you think of one or two scaled questions that would be useful for
your own research?

1 .……………………………………………………..

2 ……………………………………………………..

53
 QUESTION TYPE 6:
OPEN QUESTIONS

Allow respondents to use their own words

These are questions where the respondent can answer in their own
words. ‘Other’ is an open question. Open questions are particularly
useful for exploring parts of respondents’ experiences or
knowledge that you may not be able to foresee. For some
respondents, this can help to reduce frustration from the limiting
aspects of closed questions. For others, answering open questions can
be too much effort. And if you’re using questionnaires on paper, people
will sometimes give written information you haven’t even asked for!
Another downside is that data from open questions (or unsolicited data)
is more difficult to analyse than data from closed questions.

54
Types of open questions

Open questions can be as narrow or as broad as you like. Here are

three examples:

• Who is your favourite film star?

• What are you looking forward to most about next semester?

• Is there anything else you would like to say about the subjects
covered in this questionnaire?

Because open questions require more input from respondents and more
analytic work from researchers, they should be used sparingly.

Can you think of one or two open questions that would be useful for
your own research?

1 .……………………………………………………..

2 ……………………………………………………..

55
 COMPILING YOUR
DRAFT QUESTIONNAIRE

Bring together all the questions you have written on pages 41–55

56
 CHECKPOINT

Read through your questions. Does the order seem sensible? If not,
indicate how you would move the questions around to put them in a
more logical order.

57
 NOW YOU HAVE THE
FIRST DRAFT OF YOUR
QUESTIONNAIRE!

58
59
 You can find
people in
multiple ways

60
Section

61
62
 How do I
find people
to fill in my
questionnaires?
 10
SEC

summary

With diligence,
patience, and
politeness.

64
 60
SEC

summary

Look here, look there, look everywhere!

First, figure out what kind of people you want to ask. Your research
question may specify your participants. If it doesn’t, you will need
to define your own participant group with reference to your research
question. As a whole, this group is called a ‘population’.

Then work out how to find people from your participant group.
Generally you can find people in person, by post or email, or online. It
will be easiest if you can find people through your own personal and
professional networks, though address lists and social media are other
options. Ask politely and be prepared for some people to refuse. The
people who agree to complete your questionnaires – your respondents –
are your ‘sample’ from the population.

65
 If you wanted to research students’ experiences of managing
coursework, the population would in theory be ‘all students’. In practice
you’d need to narrow that down so you might decide the population
would be first-year students, or full-time students, or students in your
faculty or department. Then the number of students who complete your
questionnaire, from your chosen population, would be your sample.

POPULATION
AND SAMPLE

66
There is no hard-and-fast rule about how many people should
be in your sample. Occasionally, where numbers are small, you can
ask everyone in a given population to complete your questionnaire.
Otherwise you will need to use common sense. Knowledge of different
sample types can help. Here are the most relevant types of sample:

Convenience sample – people who are most likely to agree.

Stratified sample – people from each of several distinct

sub-populations. For example, with students, you might try to find
students from each year to complete your questionnaires.

Purposive sample – people who fit the particular purpose of your

research. For example, if you want to study success in student sports,
you would try to find students who are successful in sports.

SAMPLE SIZE
AND TYPE

67
 Doing this is easiest when your research is focused on a specific
place, such as investigating people’s views of a visitor attraction. It is
also easier when your potential participants are people who congregate
in accessible places. For example, if you wanted to study new mothers,
you might go to parks or cafés.

In some places you can leave questionnaires for potential participants

to find and complete. For example, if you want responses from
students, you could leave questionnaires at appropriate places on
campus: students’ union, library, gym, and so on. Staff members
in such locations might be willing to help by pointing out your
questionnaires and encouraging people to respond.

FINDING PEOPLE
IN PERSON

68
If you want to find people by post or email, first you will need to
find their postal or email addresses. It may be possible to find
these online. You can send questionnaires generically, e.g. to ‘The
Householder’ by post, or to email addresses such as info@[company].
com. However, response rates for questionnaires sent generically are
likely to be very low. Where possible, if using post or email, it is best to
send questionnaires to named individuals.

FINDING PEOPLE
IL
BY POST OR EMA

69
 The easiest way to find people online is through search engines
or social media. If your questionnaire is online, you can send people
the url via instant messaging or social media. Instant messaging
is personal so you have some control over who you send the
questionnaire to. Sharing a questionnaire link on social media can help
you get more responses, but you have no control over who answers
your questionnaire. You don’t even know for sure that they’re from
your intended participant group – remember people can conceal their
identities online.

FINDING PEOPLE
ONLINE

70
You need to tell people what you are researching and why, and how you
plan to use the data they may provide. It is essential to ask politely
and to be prepared for the answer ‘no’. If someone says ‘no’, you
should accept their refusal gracefully.

Your university’s ethical approval system may require you to give an

information sheet to potential respondents and/or ask those who agree
to sign a consent form.

If someone seems a little reluctant or uncertain, it is OK to provide

gentle reassurance, such as by reminding them that responses will be
anonymous. However, it is not OK to try to persuade or force someone
to take part in your research.

ASKING PEOPLE
T
WITH YOUR RESEO HELP
ARCH

71
 TEST YOUR APPROACH
DO IT YOURSELF

could
e se n t ences you
Write t h re ng to a
y ou a re first speaki
use when participant
potential

72
 DO IT YOURSELF

Think about this approach...

Have you properly introduced yourself?

Have you found out if they have time to spare to help?

Have you fully introduced your research?

 Have you asked if they would consider completing a

questionnaire?

e
e goals into thre
If you fit thes y on !
ces, carr
friendly senten

73
 FINDING PEOPLE

which you
t h re e ways in e
Write d ow n to complet
t ry t o find people a b ou t
could cific
e st io n n a ire. Be spe t h is .
your qu ld do
how you cou
where and

74
76
Every questionnaire
should be tested

77
78
Why do I need
to test my
questionnaire?
 10
SEC

summary

To find out whether

each question
works, and to find
out whether the
questionnaire works
as a whole.
80
 60
SEC

summary

Testing testing one two three!

However carefully you write your questionnaire, you will need to test
its questions and layout before you have a final version. This means
finding a few people to complete your draft questionnaire and give you
feedback on their experience.

Your questionnaire testers should be from your intended participant

group or as similar as possible to people in that group. As well as
asking them questions about their experience, if you can, you should
watch them work through your questionnaire. This can help you pick up
difficulties they have that they may be reluctant to admit to you or even
themselves.
81
 WHAT QUESTION
NA
TESTING MEANS IRE

Testing a questionnaire is called ‘piloting’, like testing out a new TV

series. You need to pilot your questionnaire to find out how well it
works in practice and whether you need to make any changes before
you use it for real.

When you are piloting a questionnaire, you are testing two things:

1 How well do the individual questions work?

2 How well does the questionnaire work as a whole?

Occasionally a questionnaire works well straight away. If you have

piloted it with people from your participant group, you can then use the
responses as part of your data. But mostly you will need to revise the
questionnaire as a result of your pilot investigations.

82
 WHY TEST
QUESTIONS?

It is essential to test each individual question to find out how people

will react. There may be any number of problems you haven’t foreseen,
from an un-noticed typo to unwittingly offending someone from a
different culture or religion. Piloting your questions will help you to
identify and solve such problems. Also, how you write your questions
will influence how people respond. Piloting your questions should help
you to identify and remedy any undesirable influence your question-
writing may have.
 HOW TO TEST
QUESTIONS

If you’re unsure how to word a question, you can test different versions
of that question. It is probably best to test different versions with
different people. If you decide to test different versions with the same
people, it would be useful to put them in a different order for different
testers. This is because people’s perceptions of a question may be
affected by the preceding and following questions.
 H Y T E S T T H E W HOLE
W ?
QUESTIONNAIRE

You need to test the whole questionnaire for ‘flow’ and for the
quality of its layout. What seems like a logical order to you may not
seem so to someone else. Also, is the layout you have used providing a
good experience for respondents? For example, if you’re using a paper
questionnaire, is there enough space for respondents to write their
answers? If your questionnaire is online, can respondents complete it
without spending too much time scrolling down?
 WHO TO TEST TH
QUESTIONNAIRE E
ON

You need to pilot your questionnaire with people other than those who
will be your actual participants. However, they should be as similar to
those participants as possible. If you can, pilot your questionnaire
with people from your intended participant group. For example, if
you want students from your institution to participate in your research,
then pilot your questionnaire with students from your institution. Just
make sure that when you actually collect your data, you don’t re-use
any of your testers.

If you can’t pilot your questionnaire with potential participants, try to

find people with similar attributes. What those are depends on your
research, but could mean, for example, trying to find people of a similar
age or with a similar occupation.

86
 HOW TO TEST TH
WHOLE QUESTIO E
NNAIRE

Tell people you’re testing rather than doing it for real, and ask them
to tell you about anything that’s unclear to them. Watch them while
they’re trying out your questionnaire and note any hesitation that might
indicate lack of clarity.

You can ask them questions too. For example, if they skip a question,
you can ask them why. It may help to remind them that you’re aiming
to improve your questionnaire, as some people may worry that they are
‘doing it wrong’.
 FILL IN THE
MISSING WORDS

FEEDBACK

POTENTIAL

PARTICIPANT

QUESTION

QUESTIONNAIRE

SIMILAR

TELL

TESTERS

WATCH

88
 CHECKPOINT

I need to test how well each __________________ works and how well

my __________________ works as a whole. To do this I need to find

some people from my __________________ __________________ group

or some people who are as __________________ as possible to those in

that group. I should __________________ those people that I want them

to be __________________ , ask them to give me __________________

on their experience, and __________________ them as they complete

the questionnaire.

89
90
sec
tion

6
6 Maximise your
response rate

91
92
 How can I
improve my
response
rate?
 10
SEC

summary

By following up
questionnaires,
offering give-aways or
incentives, or by making
your questionnaire
visually interesting.

94
 60
SEC

summary

How to improve your response rate

There are several things that may help improve your response rate.
While none are guaranteed to work for you, they have all been shown to
work in some circumstances.

If you have sent out questionnaires by post or email, you can follow
them up after a suitable period of time – maybe a week or two – to
remind people who may have forgotten.

If you are sending questionnaires by post or administering them in

person, you can include a small give-away such as a wrapped sweet.

In any circumstance you can include an incentive such as a chance to

win a prize draw. Or you can work to make your questionnaire more
visually appealing. And you can use more than one method if you wish.

95
 WHAT IS A
’?
‘RESPONSE RATE
The response rate is the number of people who answer your
questionnaire. Traditionally this was expressed as a proportion of
those who received the questionnaire. You can still do it like this if you
administer questionnaires in a way that involves you giving or sending
them to specific people. However, if you use social media or leave a
pile of paper questionnaires somewhere, you’ll need to use an absolute
response rate. This means simply stating how many questionnaires you
get back.

96
 HOW MANY
S
QUESTIONNAIRE
DO I NEED?

There is no hard-and-fast rule about how many questionnaires

you need, so check with your tutors and follow their guidance.
However, there are ways you can encourage people to complete your
questionnaires without being coercive. Some of these are listed in this
section. Also, if you are administering your questionnaires by post or
email, sending follow-up reminders can help to improve your response
rates. You need to word these carefully, to acknowledge that some
of the people you’re re-contacting may already have completed and
returned your questionnaire.
 OFFERING
INCENTIVES

If you have the resources, you can offer an incentive. Perhaps you have
an unwanted gift you could offer as a prize for a draw. If you have a
particular skill, such as baking, you could offer to make a cake for the
winner. Or you could offer an hour of your time, to do anything lawful,
decent, and within your abilities. There are many options – use your
imagination!

For a prize draw, you will need to give all respondents the opportunity to
enter. If they want to, they would need to give you a way of contacting
them, such as an email address or mobile number. You will need to
reassure participants that this contact information would not be
used to identify their responses.

98
 GIVE-AWAYS

Give-aways can only be used with questionnaires that are posted

or administered in person. Any give-away should be small and
inexpensive. A wrapped sweet is ideal, as the recipient can keep it
for a while or give it to someone else if they wish. If you are posting
questionnaires, choose something that will fit easily in the envelope and
not add to your postage costs. In person, give-aways should be freely
given at an early stage, including to people who express interest but
then decide not to complete your questionnaire. Give-aways should
not be used as a subtle or unsubtle form of bribery because that
would be unethical.

99
Yes


N o
MAKING
QUESTIONNAIRES MORE
VISUALLY INTERESTING

Another way to increase your response rate is to make your

questionnaire visually interesting. This can make it more attractive and
enjoyable to complete. Not everyone has the digital or artistic skills
to do this effectively. If you don’t, stick to a clear and good-sized font
and maybe introduce some colour. If you do have relevant skills, again
you can let your imagination help you. Try to include elements that will
appeal to your intended participants. Be careful, though, about using
images you find online. You should not infringe any image-maker’s
copyright by using their work without permission or payment. If you
want to source images online, use a reputable site such as Pixabay
where you can find images that are free to use.

101
 improve
What will you do to
s?
your response rate

Will you follow up questionnaires, offer give-aways or incentives, or

make your questionnaire visually interesting? Perhaps you will do more
than one of these. Maybe you have thought of something else! Write
your answers below giving full details of what you will do and how.

102
CHECKPOINT

103
 section

104
Analysing
your data
106
 What do I do
when I have all
my completed
questionnaires?
 10
SEC

summary

You prepare your

data for analysis.

108
 60
SEC

summary

Now it is time to prepare your data

There are several steps to preparing data, though if you collect data
digitally some of them will be done for you. Whatever method you
have used to collect your data, you need to check its integrity. This
means making sure each individual set of responses does not include
contradictory answers.

If you have collected paper questionnaires, you will need to enter the
responses into a spreadsheet carefully and accurately. This will also
require you to make decisions about how to handle missing data. If you
have collected your data online, via text message, or on a device, it will
need little or no further preparation.

109
 The first job is always to check through the responses from each
questionnaire in turn to make sure they are internally consistent.
Suppose you are investigating people’s views of clothing fabrics. A
respondent initially indicates that they prefer natural to synthetic fabrics.
Then, in a more detailed question, they say their three favourites are
acrylic, rayon, and viscose – all of which are synthetic fabrics. Perhaps
this is because they don’t really know which fabrics are synthetic, or
what ‘synthetic’ means. Maybe they were feeling mischievous and
decided to give you contradictory answers. Or it could be for some
other reason – and you will never know why, so you need to discard any
data from anyone who gives responses that contradict each other.

R
PREPARING YOU SIS
DATA FOR ANALY

110
You will need to number each questionnaire and enter the responses
into a spreadsheet with corresponding numbered rows. Then if you find
what looks like an anomaly in your data, you can easily check back with
the original questionnaire.

It is helpful to code your data. For example, if your participants

are other students, and you ask whether they are undergraduate or
postgraduate, you could code their responses as U/P or 1/2. This
makes entry quicker and analysis easier.

Data entry is boring but must be done carefully and accurately.

When you have finished, spot-check your work against the original
questionnaires. If you find errors you may need to check it all, which is
even more tedious than being careful in the first place.

PREPARING DATA
FROM PAPER S
QUESTIONNAIRE

111
 You will need to decide what to do with missing data. There are two
kinds of missing data: expected and unexpected. You expect missing
data in a question such as ‘Which leisure activities do you enjoy?’
followed by a list of pre-determined answers. Few if any respondents
will choose all the answers, so there will be some missing data. This is
usually dealt with in spreadsheets by leaving empty cells. Unexpected
missing data is where someone has skipped one or more questions, not
providing any answers at all. This is usually dealt with by using a code
that would not be used for anything else, such as 999 or XYZ.

DEALING WITH
MISSING DATA

112
If you have done your questionnaires online or by text message, the
data should be automatically available in spreadsheets, ready for
analysis. If you have used a device to collect your data, data should
similarly be ready for analysis. However, you will probably have to find
an internet connection before you can access the data.

Once you have finished your preparation, you’re ready to analyse your
data.

PREPARING DATA
FROM DIGITAL
QUESTIONNAIRE
S

113
CHECKPOINT

Got it?
Q: What do we need to
do when we have our
questionnaire data?

114
115
Got it!
A: E
 verything necessary to
get it ready for analysis.
 u
have mastered all yo
To help ensure you kli st
through this chec
need to know, work

Section 1 Are you confident that questionnaires will help you to

answer your research question? If not, go back to page 13.

Section 2 Do you know how you will administer questionnaires,

how they can be completed and returned, and how to ensure
participants’ anonymity? If not, go back to page 23.

Section 3 Is your draft questionnaire complete with questions

of different types that have been carefully designed to help you
answer your research question? If not, go back to page 37.

HOW TO KN
OW
YOU
ARE
DONE
116
 CHECKPOINT

Section 4 Do you know how and where to find people to fill in

your questionnaire? If not, go back to page 63.

Section 5 Do you understand why and how to test your

questionnaire? If not, go back to page 79.

Section 6 Do you know what steps to take to improve your

response rates? If not, go back to page 93.

Section 7 Do you know how to prepare your questionnaire data

for analysis? If not, go back to page 107.

117
 Glossary

Anonymity Not disclosing respondents’ identifying details, such

as their names or addresses.

Closed question A question with predefined answers for a

respondent to choose from.

Dichotomous question A question with two opposing answers

and sometimes also an intermediate answer.

Intermediate answer An answer such as ‘sometimes’ or ‘maybe’

given as an extra choice in a yes/no question or a dichotomous
question.

118
Multiple-choice question A question with several responses.

Open question A question with no predefined answer.

Participant Someone who takes part in research.

Piloting Testing a questionnaire.

Population All of the potential participants in a group you define.

119
 Respondent Someone who responds to a questionnaire.

Sample The people who participate in your research by

responding to your questionnaire.

Scaled question A question with a scaled response.

Yes/no question A question with the answers ‘yes’ and ‘no’

(a form of dichotomous question) and sometimes also an
intermediate answer.

120