LITTLE

QUICK FIX:
USE YOUR
QUESTIONNAIRE
DATA

#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:
USE YOUR
QUESTIONNAIRE
DATA

Helen
Kara
SAGE Publications Ltd © Helen Kara 2019
1 Oliver’s Yard
55 City Road First published 2019
London EC1Y 1SP
Apart from any fair dealing for the purposes of
SAGE Publications Inc. research or private study, or criticism or review, as
2455 Teller Road permitted under the Copyright, Designs and Patents
Thousand Oaks, California 91320 Act, 1988, this publication may be reproduced, stored
or transmitted in any form, or by any means, only with
SAGE Publications India Pvt Ltd the prior permission in writing of the publishers, or in
B 1/I 1 Mohan Cooperative Industrial Area the case of reprographic reproduction, in accordance
Mathura Road with the terms of licences issued by the Copyright
New Delhi 110 044 Licensing Agency. Enquiries concerning reproduction
outside those terms should be sent to the publishers.
SAGE Publications Asia-Pacific Pte Ltd
3 Church Street
#10-04 Samsung Hub
Singapore 049483

Library of Congress Control Number: 2019936583

Editor: Aly Owen
Editorial assistant: Lauren Jacobs
British Library Cataloguing in Publication data
Production editor: Ian Antcliff
Marketing manager: Ben Griffin-Sherwood
Design: Lisa Harper-Wells
Typeset by: C&M Digitals (P) Ltd, Chennai, India
A catalogue record for this book is available
Printed in the UK
from the British Library

ISBN 978-1-5264-9109-1 (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

2
 Everything in this book!������������������������������ 4

Section 1 What is data analysis?���������������������������������� 9

Section 2 What is data preparation?�������������������������� 23

Section 3 What is quantitative data coding?�������������� 37

Section 4 How can I analyse quantitative data?��������� 57

Section 5 How can I code and analyse

qualitative data?����������������������������������������� 79

Section 6 How can I interpret my findings?�������������� 101

Glossary��������������������������������������������������� 114

3
 2
MIN

summar
y

Everything in
this book!

Section 1 Data analysis is the best part of research, because this

is where you find out what your data can tell you. It is important to
analyse data carefully and honestly and to keep notes of your insights.

Section 2 Data preparation must be meticulous, even though it can

be a monotonous task. It is essential to check your prepared data to
ensure that it is as accurate as possible.

Section 3 Quantitative data is coded by assigning a number, letter,

word or phrase to each piece of data. You also need to address any
missing, contradictory, or unsolicited data.

4
Section 4 Descriptive statistics are straightforward calculations
that help you to describe and summarize your data. They include
frequencies, percentages, averages, and ranking.

Section 5 Qualitative data analysis uses all your own experience

and knowledge in a systematic way. This involves working with
codes, categories and themes.

Section 6 Interpreting your data makes your findings accessible

to others. You will create a story from your key findings that meets the
needs of your audience(s).

5
 Find out what your
data can tell you

6
Section

7
8
What Is Data
Analysis?
 10
SEC

summary

Data analysis is the exciting part

of research where you find out
what your data can tell you!

10
 60
SEC

summary

One step at a time

The process of analysing data includes several steps. You have to:

• Prepare your data

• Code your data
• Analyse your data to produce your findings, and
• Interpret those findings for your readers.

Fortunately you can take this one step at a time. This book is
structured to help you.

Data analysis operates differently for quantitative and qualitative data.

Whatever kind of data you have, it is important to analyse it ethically,
i.e., carefully and accurately. You will need to keep notes of your
findings and thoughts as you analyse your data. These notes will help
you to produce your report or assignment.
11
 WHAT IS DATA
ANALYSIS?
Your data is the responses from your questionnaires. Analysing data
is a process with several steps: preparation, coding, analysis itself and
interpretation. Parts of the process operate differently for quantitative
data (made from numbers or from responses that can be counted) and
qualitative data (made from words).

All data analysis, whether qualitative or quantitative, has ethical aspects.

Also, data analysis is closely linked with writing.

This may all sound very complex, but you can take it one step at a time.

12
 WHAT KINDS OF
DATA DO I HAVE?

That depends on what kinds of questions you asked in your

questionnaire. Most questions will generate quantitative data. These
include, for example:

• Demographic questions that generate information about your

participants, such as their gender

• Yes/no questions that classify those people into groups, such as

those who like football

• Dichotomous, multiple-choice or scaled questions to find out about

people’s experiences.

Questions that generate qualitative data come in two categories.

One is a box marked ‘other’, or its equivalent, which can be added
to a multiple-choice question. The second is a straightforward open
question, for instance, ‘please tell me about your experience in your
own words’.

13
 ISN’T DATA
ANALYSIS
REALLY HARD?

Some researchers find data analysis quite difficult, while others think it
is the easiest part of research. Some people find quantitative analysis
easy and qualitative analysis difficult; others are the other way around.
Everyone is different and you won’t know whether data analysis is hard
for you until you give it a try.

This book sets out clear procedures for using your questionnaire data.
Your best chance of finding data analysis as easy as possible is to
follow the procedures carefully, step by step.
 THE ETHICS OF
DATA ANALYSIS

Researchers generally analyse data alone. Even so, they may be aware
of all the people who have an interest in their findings: their participants,
tutors, families and friends. Also researchers have their own biases and
pressures. I think it is a rare researcher who has never been tempted
to tweak their data or their findings to fit their own or someone else’s
agenda. Yet doing this is highly unethical.

The most ethical approach to data analysis is to analyse all your

data thoroughly and fairly. If you follow the procedures set out in this
book you will be able to analyse your data ethically.

15
 IS
DATA ANALYNSG
AND WRITI
Writing is inextricably linked with data analysis; you can’t do one and
then the other. As you analyse your data, you need to write down what
you find out, as well as any thoughts that arise about your research
question. This writing doesn’t need to be neat. It could be on sticky
notes, electronic memos, even voice clips on your phone: whatever
works best for you. However, it is a good idea to keep all your writing in
one place, and to make sure it’s legible (or audible). Then you will find it
easier to use your analytic notes to help construct your assignment or
report.

16
17
CHECKPOINT

Get it?
Q: How can I be sure my
data analysis will be
ethical?
Got it!
A: Analyse your data
thoroughly and fairly.
Write down what you find
out, and don’t tweak your
findings!
 Section
20
 Data preparation
must be meticulous

21
 What
Is Data
Preparation?
 10
SEC

summary

A way of getting your data ready

for coding and analysis.

24
 60
SEC

summary

Accuracy is paramount

Data from questionnaires administered electronically should arrive

ready-prepared. However, you would still need to check, as best you
can, that the data seems accurate – or at least plausible.

Data from paper questionnaires needs preparation. You need to

number each questionnaire consecutively and then enter the data from
each questionnaire into one row of a spreadsheet.

It is crucial to ensure that your data is as accurate as possible. This

means you also need to check the accuracy of the data entry, even if
you have done that yourself.

The whole process of preparing data is quite tedious so is best done in

small chunks in between other tasks.

25
 HOW CAN
I PREPARE
MY DATA?
How you prepare your data for coding depends on how you
administered your questionnaire. The most common methods are
on paper, online, by SMS, or using an electronic device such as a
tablet. Data from questionnaires administered electronically should be
prepared by the software you use, which should provide the data in a
format such as a ready-to-code spreadsheet.

Conversely, data from paper questionnaires will require quite a lot

of preparation. The first thing to do is number the questionnaires in
writing, consecutively, starting with ‘1’. Then you need to enter the
data into a spreadsheet, one row per questionnaire, beginning with the
questionnaire number.

26
 NTER
HOW DO I EA
THE DAT ?
Entering quantitative data will involve making some initial coding
decisions. For example, suppose you asked, ‘Which gender do you
identify with most?’ and gave the answer options:

1 Male
2 Female
3 Both
4 Either
5 Neither

You could enter the initial letters M, F, B, E and N into your spreadsheet,
or you could use the numbers 1–5. It doesn’t matter which; they both
enable sorting and counting, which is useful at the analysis stage.

Alternatively, imagine you have a scaled question where you ask

people to respond with a number between 1 and 7. You can use those
response numbers as your codes.

You will need to enter any qualitative data as it stands for coding later
on. (See Section 5.)

27
 AKE
WHAT IF I ME
A MISTAK ?
It is essential to ensure that your data is accurate. This means
being meticulous – and that can be very boring. It’s important, though,
because a single error could completely change your findings, such as if
you put a zero or a decimal point in the wrong place.

There are various kinds of errors you could make, such as:

• An error of omission, where you leave something out

• An error of commission, where you put something in that shouldn’t

be there

• A transposition error, where you put two numbers or words in the

wrong order

• A typographical error, where you intend to type one thing but in fact
type another.

28
 H O W W O U L D I
W I F I ’ D M A D E
KNO
A MISTAKE?
You need to look through your prepared data to check that it seems
plausible. Data collected online is most likely to be accurate, but
mistakes can still creep in. (In fact, with this type of data, you may never
know for sure whether it is accurate!)

Contradictory data within one participant’s response is a red flag (see

next section for more on this). If you see something that surprises you,
such as everyone giving the same answer to one question, think about
the possible reasons. Could it be that all your participants truly had the
same opinion or experience, or might it indicate an error?

Even if you have entered the data yourself and been very careful, you
still need to check it for mistakes.
 IS DATA PREPARATION
AS BORING AS IT
SOUNDS?

Honestly? It can be.

The best way to manage the monotony involved in data preparation

is to prepare your data a little at a time. Ideally you would do this
while you’re still collecting your data. However, if you have all your
data already, then break it up into chunks to prepare – say, three
questionnaires to enter, or two sets of electronic answers to check –
and do other things in between.

30
31
 CHECKPOINT

Fill in the
missing words

Using the words below, fill in the gaps in these sentences.

Data preparation can make you feel quite ………… but it is

important to work ………….. You need to do all you can to avoid

making ………… and this involves …………. to make sure your data

seems ………….. A good way to reduce boredom is to …………..

your data a ………… at a time.

accurate checking
bored little
carefully mistakes
prepare

32
33
ANSWERS
Data preparation can make you feel quite …………
bored but it
is important to work carefully. You need to do all you can to avoid
making mistakes and this involves checking to make sure your data
seems accurate. A good way to reduce boredom is to prepare your
data a …………
little at a time.
 Section

34
 Assign a number,
letter, word or phrase
to each piece of data

35
 What Is
Quantitative
Data Coding?
 10
SEC

summary

A systematic way of ensuring

that each piece of quantitative
data has an appropriate code.

38
 60
SEC

summary

Number, letter, word or phrase

Each quantitative response to a question needs to be given a code. This

code may be a number, a letter, a word or a phrase. Some codes are
assigned during data preparation, but all data must be coded before
you can start work on your analysis.

Some data can be auto-coded, which means you can choose the
response itself to become the code. This can be done with questions
that have numerical answers, such as scaled questions.

You will need to assign a code to any piece of data that cannot be auto-
coded. You will also need to assign a special code to missing data and
know how to deal with contradictory or unsolicited data.
39
WHY ARE QUANT
AND QUALITATIV ITATIVE
E
CODED DIFFEREN DATA
TLY?

Coding involves assigning a code – a number, letter, word or

phrase – to a piece of data. As we have seen, if you have paper
questionnaires, you will start coding quantitative data as you input your
data on to a spreadsheet. If your questionnaire was online, via SMS, or
using an electronic device such as a tablet, much of the coding may be
done for you. However, it is still important to understand the principles
of data coding so that you can check it has been done correctly.

Conversely, qualitative data is entered as it stands, for coding at a

later date. Section 5 explains how to code qualitative data.
 WHAT IS
AUTO-CODING?

Some data can be auto-coded. This doesn’t mean it codes itself, it

means that the response becomes the code. This applies to questions
made from numbers, such as scaled questions like this:

How confident do you feel about finding

a job after finishing university?

Totally confident 7 6 5 4 3 2 1 Not at all confident

You can see that participants’ responses can be used, in themselves, as

the codes.

41
 WHEN AND HOW
DO I ASSIGN
CODES TO DATA?

When answers cannot be auto-coded you need to assign codes. For

example, if your participants are university students, you might ask
whether they are studying for an undergraduate degree, a Master’s
degree or equivalent, or a doctoral qualification. You could code their
answers using letters – perhaps the initials U, M, D – or numbers,
i.e., 1, 2, 3. Either system will enable you to input swiftly and sort your
spreadsheet later.

So far, so straightforward. However, you also need to deal with missing

data, contradictory data, and perhaps unsolicited data.

42
 HOW CAN I DEAL
WITH MISSING DA
TA?

Missing data comes in two categories. Some is caused by

participants skipping questions, whether by accident or on purpose.
Other missing data is due to question structure, such as multiple
choice. Look at this question:

Which housework activities do you enjoy? Please tick all that apply.

Dusting Cleaning the toilet

Vacuuming Cleaning out the fridge

Ironing Cleaning the oven

Cleaning the kitchen Making beds

Cleaning the bathroom Other

A few people may tick all of these, but most will tick some or maybe
only one. This means that the others will be classed as ‘missing data’.

A good approach to missing data is to use one code for data from
skipped questions and another for data from multiple-choice questions.
They should be codes you couldn’t possibly use for anything else, such
as XXX or 9999.

43
 WHAT ABOUT
CONTRADICTORY
DATA?

Contradictory data within a single response presents a problem.

Consider these questions and their answers.

How satisfied are you with the teaching of this module?

Perfectly satisfied 7 6 5 4 3 2 1 Not at all satisfied

Can you tell us why?

My teacher was great, I learned loads

44
The respondent has said they are not at all satisfied with the teaching
because it was really good. This is contradictory, and you don’t know
why. The respondent could have made a mistake, or it could have been
a mischievous prank. Maybe there was some other reason; you will
never know. Therefore you need to remove their data from your dataset
because it is unreliable. This can be a blow, particularly if you have
found it difficult to recruit respondents, but it is the only ethical course
of action.
 D
AND UNSOLICITE
DATA?

Unsolicited data is where a respondent provides information you

have not asked for. They may do this by writing in the margins of a
paper questionnaire or in a text box in an electronic questionnaire. If the
information provided is relevant, you can include it if you like, as long as
you make its status clear by explaining how it was provided.

Paper questionnaires can yield other forms of unsolicited data. For

example, someone faced with the scaled question above, about
how satisfied they are with the teaching of their module, might circle
numbers 3 and 4, or write X in between the two. You could code this
as 3.5, or you could decide that you don’t understand their meaning
so you will code it as missing data. Either is fine as long as you do the
same thing for each instance (if there is more than one).

46
47
 HOW TO CODE
DO IT YOURSELF QUANTITATIVE DATA

How would you code the answers to the following questions?

1 What type of place do you live in? Please tick one.

Own house Parents’ flat

Parents’ house Shared flat

Shared house University residence

Own flat Other

OH, PH, SH, OF, PF, SF, UR, O

1 assign a number from 1–8 to each answer, or use initials i.e.,

Answer

48
 DO IT YOURSELF

2 How confident do you feel about the quality of your housework?

Very confident 7 6 5 4 3 2 1 Not confident at all

1 use auto-coding i.e., the numbers 1–7

Answer

49
 HOW FAR HAVE
YOU COME?

Think about the questions below and tick any you can honestly answer
with a ‘yes’. If there are any where you don’t feel able to answer with a
‘yes’, check back to see what you still need to do.

Have you prepared your data carefully?

 ave you checked to make sure your prepared data is as accurate

H
as possible?

Have you coded all of your quantitative data?

Have you dealt with any missing, contradictory or unsolicited data?

50
CHECKPOINT

51
52
Congratulations!
You have prepared
and coded all of your
quantitative data!

53
 Describe and analyse
quantitative data

54
Section

55
 How can
I analyse
quantitative
data?
 10
SEC

summary

You can analyse quantitative data

using descriptive statistics.

58
 60
SEC

summary

Describe and summarise

Descriptive statistics help you to describe and summarise your

data. They include frequencies, percentages, averages and ranking.
Frequencies answer the question ‘How many?’; percentages answer the
question ‘What proportion?’; averages help to show representation, and
ranking puts your data in order.

Some people feel anxious about statistics, fearing they may be too
difficult. It’s true that some statistical calculations can be complex but
descriptive statistics are straightforward. This section explains when
to use frequencies, percentages, averages and ranking, and how to
work them out. If you follow the instructions carefully you will soon be a
descriptive statistics wizard.
59
 S
AREN’T STATISTIC
REALLY HARD?
Statistics can be very complicated but they don’t have to be. The more
complex statistical calculations can only be used with data from a truly
random sample. As a student researcher, you are more likely to use a
different kind of sample. This might be a:

Convenience sample – people who you could easily get to agree to

help you

Purposive sample – people who fit the purpose of your research. For
example, if you were studying people’s opinions of cinema facilities, you
would need people who visited cinemas

Stratified sample – people from each of several sub-populations. For

example, if you wanted student participants, you might want people
from each faculty or department.

With these kinds of samples you can use descriptive statistics.

60
 WHAT ARE
DESCRIPTIVE
STATISTICS?

Descriptive statistics summarise and describe data for your

readers. These statistics include frequencies, percentages, averages,
and ranking. If you’re not familiar with those terms this may seem
daunting, but don’t worry, they’re really straightforward.

Frequencies are about counting

Percentages are about proportions

Averages are about representation

Ranking is about ordering.

The key is to know two things:

1 What each statistic is for

2 How to calculate each statistic

61
 Here’s a little dataset we can work with. It is answers to the question,
‘What is the maximum number of minutes in a day you would spend
vacuum-cleaning your home?’ Respondents were able to type in a
whole number of their choice.

1 5 8 5 15 2
2 7 9 8 16 12
3 10 10 10 17 10
4 5 11 5 18 15
5 20 12 15 19 10
6 0 13 5 20 0
7 3 14 8

62
FREQUENCIES

To calculate frequencies, count how often each answer appears.

Frequency data is often presented in a table. The data above would
look like this:

Response Frequency
0 2
2 1
3 1
5 5
7 1
8 2
10 4
12 1
15 2
20 1

If you make the table using spreadsheet software such as Excel, you
can convert it into a graph or chart, which is useful for presenting your
data.

63
 PERCENTAGES

The word ‘percent’ comes from the Latin ‘per centum’ meaning ‘of
the hundred’. So if you hear that 25% of people like cabbage, you
know that means twenty-five of every hundred people. Percentages
represent proportions and can often be simplified. Twenty-five of
every hundred people is challenging to visualize, but you can simplify it
to one in four, which is the same proportion.

To calculate percentages, divide the number of times a response

is given by the total number of respondents, then multiply by 100.
From the table above, you can see that four people said they spent
10 minutes vacuum-cleaning their home. Dividing 4 by 20 yields 0.2;
multiply by 100 and you have 20 percent.

64
Proportions can be useful for describing your data, not least because
they are linked to probability. However, it is important not to use
percentages to make big claims from your research. If you have
only heard from 20 people, you can’t claim that 20 percent of all people
spend 10 minutes a day vacuum-cleaning their home. You can only
claim that this is the case for 20 percent of your sample.
 AVERAGES

There are three kinds of average:

Mean
(common average)

Median
(central value)

Mode
(value found most often)

66
MEAN

To calculate the mean, add all the numbers then divide by how many
numbers you have. For our dataset, this would look like:

(5+7+10+5+20+0+3+5+8+10+5+15+5+8+2+12+10+15+10+0)
÷ 20 = 7.75

This suggests that the average person from your sample spends 7.75
minutes a day vacuum-cleaning their home. This type of average is
most useful where there are no outliers, or unusual responses, as with
our dataset.
 MEDIAN

To calculate the median, put all the numbers in order and choose the
central one. Here are our numbers in order:

0 0 2 3 5 5 5 5 5 7 8 8 10 10 10 10 12 15 15 20

As we have an even number of responses, we take the two in the centre

and calculate the mean of those:

(7 + 8 ) ÷ 2 = 7.5

With this dataset, the median and the mean are close together. The
median is most useful where there are outliers. For example, imagine
the respondent who said they spent 20 minutes a day vacuum-cleaning
had said instead they spent 180 minutes doing so. That would change
the mean to 15.75, while the median would stay the same. So where
there are a small number of outliers the median can be a more reliable
average.

68
MODE

To calculate the mode, see which response is given most frequently.

There’s a clue – we can use the frequency table. This tells us that five
people said they vacuum-cleaned their home for five minutes a day. So
the mode for this dataset is 5.

With our dataset the mode is not particularly useful. However, imagine
that 16 respondents had each said they spent five minutes a day
vacuum-cleaning. In this case the mode would be a helpful way to
describe the data.

69
 RANKING

Frequencies, percentages and averages are calculated from your raw

data, i.e., your data as it is before analysis. You can rank your raw data
too, but it is often more helpful to rank your frequencies, as in the table
below.

Frequency Response
5 5
4 10
2 0
2 8
2 15
1 2
1 3
1 7
1 12
1 20

70
With the frequencies ranked, it is easier to make sense of the data.
You may also notice that I have ranked responses that were given with
the same frequency. Two people each responded with 0, 8 and 15, so I
have put those in ascending order. They could equally be in descending
order, though it is important to be consistent: use one approach or the
other, not both. Therefore, I have also ranked the five responses given
by one person each in ascending order.

71
 HOW TO CALCULATE
DO IT YOURSELF DESCRIPTIVE STATISTICS

Here is the dataset from another question in the housework study:

‘How many times would you usually clean your oven in a month?’

1 1 8 3 15 2
2 4 9 1 16 30
3 0 10 2 17 4
4 1 11 1 18 1
5 2 12 4 19 4
6 0 13 1 20 1
7 1 14 0

Complete the frequency table for this data.

Response Frequency

72
 DO IT YOURSELF

Calculate the mean for the data.

Calculate the median for the data.

Calculate the mode for the data.

Which of these averages would you use, and why?

73
 74
ANSWERS DO IT YOURSELF
Check your
Answers
Response Frequency
0 3
1 8
2 3
3 1
4 4
30 1
Mean: 3.75
Median: 0 0 0 1 1 1 1 1 1 1 1 2 2 2 3 4 4 4 4 30 = 1
Mode: 1
The median and the mode are the same so you could use either or both.
The mean is affected by the outlier i.e. the one person who says they
clean their oven 30 times each month.
 USING DESCRIPTIVE
STATISTICS TO ANALYSE
QUANTITATIVE DATA

Do you remember what the different types of statistical calculation are

for?

Their functions are given below; can you match each one with the
correct type of calculation?

Representation
Ordering
Counting
Proportions

Frequencies are about

Percentages are about

Averages are about

Ranking is about

ranking – ordering.
Answer: frequencies – counting; percentages – proportions; averages – representation;

75
 Section

76
 Use your own
experience and
knowledge in a
systematic way

77
How can I
code and
analyse
qualitative
data?
 10
SEC

summary

By using your own experience and

knowledge in a systematic way.

80
 60
SEC

summary

Wielding the words

Qualitative data coding and analysis is all about words. The data
is made up of words and coded with words and phrases. Analysis
involves rearranging data and codes and writing new words to record
insights you experience and conclusions you draw.

While the process varies depending on how much data you have, it is
always systematic and methodical. If you have only a little data you
can skip the coding and go straight to creating categories for analysis.
If you have a lot of data you will need to code it before you can create
categories and identify themes.

81
 WHAT’S THE
H
DIFFERENCE WITA?
QUALITATIVE DAT

Methods of coding and analysing qualitative data are not specified as

precisely as for quantitative data. With qualitative data, you need to
use your own experience and knowledge. This may sound a bit vague
but you do need to be just as systematic and methodical as with
quantitative data.

Qualitative data analysis involves coding and looking for categories

and themes. How you do this depends partly on how much data you
have. If you have just a few short answers to one question, there is no
point in coding the data.

82
 HOW CAN I
ANALYSE DATA
?
I HAVEN’T CODED

You can group the answers into categories. If you spot any answers
that say the same kind of thing, group them together. If you are working
electronically you can copy each piece of data into a list with others that
say the same or something similar. If you are working from hard copies,
you can write each piece of data on a separate piece of paper or sticky-
note and put it with others that say the same, or something similar. Each
group then forms a category. It is a good idea to name each category
for ease of reference.

83
 If your qualitative data proves impossible to group, don’t panic. This
is a finding in itself. Think about what it might mean in the context of
your research. Could it reflect the diversity of your participants? Is it
in response to a question that is likely to elicit a wide range of views?
Maybe it’s because you have a small number of responses. Or is there
some other reason? Write down whatever you conclude and any other
insights.

RE NO
WHAT IF THERE ADATA?
GROUPS IN MY

84
If you have more data than just a few short answers you will probably
need to code your data. There are no hard-and-fast rules about how
much data you need to use coding. A useful rule of thumb is that if
your data feels like too much to group into categories straight
away, you probably need to use coding. Another is that if you only
have short answers, even if there are a lot of them, you can probably go
straight to grouping. Conversely, if your respondents have written whole
paragraphs you will probably need to code that data.

E N S H O U L D I C ODE
WH A?
QUALITATIVE DAT

85
 To code qualitative data, you need to read it carefully, one line at a
time. The easiest way to do this is to gather all the responses to each
question. Put them into a new worksheet or document and then read
them through. As you read, code whatever you perceive within the data
with any words or phrases you choose. Highlight each piece of data and
attach the code as a comment in the margin. Keep a list of the codes
you create and re-use them when appropriate.

This process can feel difficult at first but keep going and, as you build
your list of codes, it will get easier. When you have read through once,
read through your codes and your data again to ensure you haven’t
missed anything.

HOW DO I CODE
QUALITATIVE DAT
A?

86
There are two main ways of analysing coded qualitative data: content
analysis and thematic analysis. Which you use will depend on your
research question and the quantity of data you have. In some cases you
might use both.

HOW CAN I ANAL

Y
DATA WHEN IT HASE
S
BEEN CODED?

87
 CONTENT
ANALYSIS

Content analysis works well for questionnaire data because it is a way

of quantifying qualitative data. At its most basic form, you simply
count, although there are some decisions to be made. A key decision is
whether you count how many respondents mention something, or how
many mentions are made overall, or both.

In the housework research mentioned earlier there are 20 respondents.

Twelve of them mentioned ironing, and the total number of mentions
was 29, because eight respondents mentioned it more than once; some
several times. So there may be a big difference between the number of
respondents who make a mention and the total number of mentions.

There are no hard-and-fast rules about whether you should count how
many respondents mention something or the total number of mentions.
You may need to do both before you decide which is the best measure
to report.

88
 THEMATIC
ANALYSIS
Thematic analysis is the most thorough form of analysis but you can
only do this if you have quite a lot of qualitative data. A theme is an
idea or concept. Themes are the top or broadest level of analysis,
with categories in the middle and codes as the first level. This means
you will have more codes than categories and more categories than
themes. A theme may be derived from categories (alone or combined),
or from relevant literature you have reviewed, or from a combination of
categories and literature.

To identify categories and themes, re-read the coded data. Ask yourself
questions such as:

• Do any of these codes belong together in categories?

• What thoughts or insights occur to you?

• Do you see anything surprising?

• Can you spot any possible themes?

seems
d ow n an y thing that
Write ugh.
va n t as y ou read thro
rele

89
 HOW TO IDENTIFY
DO IT YOURSELF
CATEGORIES
AND THEMES IN
QUALITATIVE DATA
Here’s an example of some qualitative data in response to the question
‘What is your overall view of housework?’ How would you group these
answers into categories?

I like ironing and making beds, but I don’t like the rest.

I don’t do housework. It’s women’s work.

It has to be done; there’s no choice.

It’s soul-destroying because it’s never finished.

I like when my Dad does it because the house gets clean and tidy.

I don’t mind housework except for cleaning the toilet.

Hate it.

Housework is something other people do!

Resigned. It’s a fact of life.

I only enjoy sweeping floors (yeah, I know; weird).

90
 DO IT YOURSELF

Remember there are no hard-and-fast rules about how to do this.

You need to use your own perceptions and intelligence to group the
answers. Have a go at identifying some categories and supporting
statements here:

estion s on
e h av e g iv en our sugg
W ge.
the next pa

91
DO IT YOURSELF

Suggested categories:

Like some tasks, dislike others

I like ironing and making beds, but I don’t like the rest.
I don’t mind housework except for cleaning the toilet.
I only enjoy sweeping floors (yeah, I know; weird)

Somebody else’s job

I don’t do housework. It’s women’s work.

I like when my Dad does it because the house gets clean and tidy.
Housework is something other people do!

Dislike

It’s soul-destroying because it’s never finished.

Hate it.

Not optional

It has to be done; there’s no choice.

Resigned. It’s a fact of life.

92
 DO IT YOURSELF

Can you see any themes developing? Identify some possible themes
here.

estions on
e h av e m ad e some sugg
W ge.
the next pa

93
DO IT YOURSELF

Suggested themes:

Somebody else’s job

In this case a category becomes a theme.

Emotion

This runs through all of the categories; particularly notable as the

question was about ‘views’ rather than ‘feelings’. It seems that
housework is an emotional topic which may point to an interesting
finding.

This is only the initial analysis of a few answers to a single question so

it is too early to draw any conclusions. The point of this exercise was
to give you an idea of how the coding and analysis process works with
qualitative data.

94
 MATCH THEM UP

Which definitions match which word?

A code is

A category is

A theme is

Definitions:

1 A group of items of data

2 A central idea arising from data

3 A number, letter, word or phrase

assigned to a piece of data

95
 96
ANSWERS
A code is a number, letter, word or phrase assigned to a
piece of data.
A category is a group of items of data.
A theme is a central idea arising from data.
Congratulations!
equipped
You are now fully
your
to code and analyse
qualitative data!

#LittleQuickFix
97
98
Make your findings
accessible to others

6
Sec
tion
6 99
 How can I
interpret my
findings?
 10
SEC

summary

By telling the story your data

would tell if it could talk.

102
 60
SEC

summary

The beating heart of research

Interpretation is the heart of research work because it is the way

in which we make our findings accessible to others, which increases
the impact of our research work. Interpretation involves thinking,
decision-making, writing and perhaps a little extra fact-finding. The act
of interpretation requires the researcher to look in two directions: at the
story the data is trying to tell, and at the needs of the audience(s) for
your research.

A central part of interpretation is to identify your key findings and

use them to construct a narrative. While you’re doing this, you may
realize that supplementary information could be useful, in which case
you might decide to look something up online.
103
 WHAT DO YOU
MEAN BY ‘INTER
PRE
MY FINDINGS’? T

You need to use your findings to find the story your data would tell if it
could speak. This is the act of interpretation, and it involves:

• Thinking about how your findings relate to each other

• Thinking about how your findings relate to the literature you have
reviewed

• Identifying your key findings

• Working out how to weave your findings and the literature into a
narrative

• Doing any extra fact-finding that may be needed.

Interpretation is at the heart of research work because it is the way we

make our findings accessible to other people. So you need to think not
only about the story your data wants to tell, but also about the story
your audience(s) want to hear.

104
 HOW DO I IDENTIF
MY KEY FINDING Y
S?
This involves considering each of your findings in turn in the light of
your other findings and the literature you have reviewed. Not all of your
findings will be noteworthy. However, be sure to note:

• Anything that surprises you

• Any strong findings

• Any inconclusive findings

• Any relationships between different findings

• Any relationships between a finding and the literature you have

reviewed.

Strong findings are not necessarily key findings, though they may be.
A strong finding that you did not expect is more likely to be a key
finding. As so often with research, there are no hard-and-fast rules
about what is and what isn’t a key finding. It’s a matter of judgement
and the final judgement is yours.

105
 HOW DO I CREAT
E
A NARRATIVE?

Once you have identified your key findings, write each one out as a
sentence. If you are writing by hand, it will help if you write each finding
on a separate piece of paper or sticky note. Then think about the
order of your findings. Can you arrange them into some kind of logical
progression? Examples include:

• chronological order

• small to large

• first to last.

If there is no obvious logical progression, think about your audience(s).

What will be the best place to start to help them understand your work?
Where will be the best place to go from there?

There is unlikely to be one single best way to order your findings, so

you’re looking for a useful, credible approach.

106
 WHAT DO YOU M
E
I MIGHT HAVE TO AN,
MORE RESEARCHDO
?
When you’re interpreting your findings, other questions may occur to
you. It is worth noting down these questions for two reasons. First,
you may be able to answer them quickly by doing a little fact-finding.
Second, they might lead to ideas for further research to mention in your
assignment.

The kinds of questions you might be able to answer quickly include:

• How does this finding compare with national figures?

• Has anyone done research on this topic in a comparable setting?

• Has anyone done research on a parallel topic?

These kinds of questions can often be answered in minutes by

searching online. If you search and find the answer is not available, it is
still worth noting down what you have done.
 WHY IS THIS
IMPORTANT?

If you interpret your findings thoroughly, and meet the needs of your
audience(s), you give your research the best opportunity to make a
difference. You may feel that your student research project has no hope
of creating change but some truly do. Even if it doesn’t – even if you
hate doing research and can’t wait to finish – it is still worth doing your
project as well as you can. There are two reasons for this.

First, obviously, you will get better marks for thorough rather than
slapdash work. Second, you will inevitably use research during your
career. You will be better placed to assess the quality of research if you
have done a good piece of work yourself.

108
WHAT DO I DO NEXT?

Your next job is to write your assignment! Some people are confident
writers, others less so. If you feel you need help with writing, check
out The Quick Fix Guide to Academic Writing by Phillip C. Shon, or
Writing Successful Reports and Dissertations by Lucinda Becker.
These books cover the most common questions and problems students
have about writing.
CHECKPOINT

Get it?
Q: How do I interpret
my findings?

Identify your key A:

Got it!
A: Identify your key
findings and use them
to create a narrative that
will meet the needs of
your audience(s).
 Can you agree with
ements?
each of these stat

Have you:

Understood the process of data analysis, its ethical aspects, and its
relationship with writing? If not, go back to page 11.

Prepared your data carefully and checked it for accuracy? If not, go

back to page 26.

Coded your quantitative data and, if you have enough of it, your
qualitative data? If not, go back to page 40 for quantitative data
and 81 for qualitative data.

HOW TO KN
OW
YOU
ARE
DONE
112
 CHECKPOINT

Used descriptive statistics to describe and summarise your

quantitative data? If not, go back to page 59.

Worked to produce categories, and perhaps themes, from your

qualitative data? If not, go back to page 82.

Identified your key findings and used them to tell a story? If not, go
back to page 104.

If you can agree with each statement – then you are done!

113
 Glossary
Analysis a systematic way of working with your data to
produce findings.

Average a number expressing a central or typical value from

a set of quantitative data – this may be the mean, median or
mode.

Category a group of pieces of qualitative data, or codes,

which have something in common; the middle level of analysis.

Code a label assigned to a piece of data; the first level of

analysis.

Coding the process of assigning labels to pieces of data.

Contradictory data data that is inconsistent with other data

from the same respondent.

Convenience sample people who you could easily get to help

you.

114
Data the responses from your questionnaires.

Descriptive statistics calculations that help you describe and

summarize your quantitative data.

Frequency counting how many people gave each response.

Interpretation telling the story of your findings.

Mean the most common calculation of an average, worked

out by adding all the numbers then dividing them by the number
of numbers.

Median the central value of your quantitative data when put in

numerical order.

Missing data data that is not present, whether through

omission or in response to a question where some answer
options may not be selected.

Mode the value found most often in your quantitative data.

115
 Percentage proportion of one hundred.

Preparation getting data ready to be coded.

Purposive sample people who fit the purpose of your

research, e.g., if you are studying the experience of playing
football you would select people who play football.

Qualitative data data made from words.

Quantitative data data made from numbers.

Random sample a sample in which every member has an

equal chance of selection.

Ranking putting your data or your frequencies in ascending or

descending numerical order.

Raw data data as it is before analysis.

Sample The people who participate in your research by

responding to your questionnaire.

Stratified sample people from each of several sub-

populations, e.g., if you are studying employees you might
choose equal numbers of part-time and full-time workers.

Theme an idea or concept arising from your data analysis and

literature review.

Unsolicited data data that doesn’t correspond to a question

you have asked or an answer option you have provided.

116