SAGE Research Methods

Grounded Theory for Qualitative Research: A Practical

Guide

Author: Cathy Urquhart

Pub. Date: 2017
Product: SAGE Research Methods
DOI: https://dx.doi.org/10.4135/9781526402196
Methods: Grounded theory, Theory, Coding
Disciplines: Anthropology, Business and Management, Criminology and Criminal Justice, Communication
and Media Studies, Education, Health, History, Marketing, Nursing, Political Science and International
Relations, Psychology, Social Policy and Public Policy, Social Work, Sociology
Access Date: November 29, 2022
Publishing Company: SAGE Publications, Ltd
City: 55 City Road
Online ISBN: 9781526402196

© 2017 SAGE Publications, Ltd All Rights Reserved.

SAGE SAGE Research Methods
© Cathy Urquhart 2013

Grounded Theory Method (GTM)

This chapter:

• explains how grounded theory method (GTM) started in 1967

• explains the key characteristics of GTM
• discusses how GTM has evolved into several strands
• discusses different coding procedures used in GTM
• discusses how various myths have evolved surrounding the use of GTM.

Introduction
This chapter aims to equip you with a thorough overview of grounded theory method (GTM). By the end of
this chapter, you should have a good understanding of its key characteristics, its intellectual history and its
diverse coding procedures. Because GTM does have such a complex intellectual history, it is necessary to
know about it, as it is definitely what Bryant and Charmaz (2007) rightly characterise as a contested tradition.
The chapter concludes with some myths about GTM and how to defend your use of the method. Let's now
start at the beginning.

The discovery of grounded theory

GTM started with a revolutionary book in revolutionary times. In 1967, Barney Glaser and Anselm Strauss
published a book entitled The Discovery of Grounded Theory. This book outlined a research methodology that
aimed at systematically deriving theories of human behaviour from empirical data.

It was a reaction against the use of ‘armchair’ functionalist theories in sociology. Glaser and Strauss (1967:
10) claimed there was a trend afoot in sociology, that the ‘great men’ of sociology had generated enough out-
standing theories and all there was left to do was test them. They further charged that the ‘great men’ played
‘theoretical capitalist’ to a mass of ‘proletariat’ testers, so sociologists were trained to only test, not imitate.
The book is a strong call to generate and ground theory and refocus on qualitative data rather than quantita-
tive verification of theories.

Many people find it difficult to read the original 1967 book, which is a pity. Melia (1996) says it has some
‘near mystical passages’, which is true. The book has to be seen in the historical context in which it was writ-
ten, which was in a decade where many new, ground-breaking ideas emerged and changed society forever.
I would highly recommend reading this book at some point during your use of GTM. It is indeed a classic

Grounded Theory for Qualitative Research: A Practical Guide

Page 2 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

and it is always good scholarship to return to the original source of ideas. I have found myself turning back
to it again and again while writing this book. No elaborate procedures are given and there is a focus on what
theory actually is, as well as the means of developing it. It also provides the baseline for understanding the
many debates that have arisen around GTM.

Several more books and articles by the co-originators followed, which developed, and later debated, the
method. Glaser's Theoretical Sensitivity, published in 1978, introduced several key concepts that are useful
in GTM. First, he talked about the role of the literature and induction. The need to be theoretically sensitive
was explained as the need to understand theories and how they are constructed, but without then imposing
those concepts on the emergent theory. He also introduced the notion of ‘coding families’ to help with relating
concepts in the data.

In 1990, Basics of Qualitative Research, by Anslem Strauss and Juliet Corbin, was published. A long and
bitter dispute erupted between Glaser and Strauss and what was at stake was nothing less than the identity
of GTM. It marked the start of what are effectively two strands of GTM and these endure to this day. We shall
return to the dispute, and why it was so important, later in this chapter.

Following the publication of the seminal work in 1967, GTM spread fairly quickly as a qualitative research
method within the social sciences and many other fields. For example, Benoliel (1996) says there was a
70-fold increase in published papers with ‘grounded theory’ as a keyword phrase in the health field over the
previous decade. By the mid-1990s, the methodological procedures of GTM had permeated qualitative re-
search to such an extent that Miles and Huberman (1994) labelled it a ‘common feature [of qualitative] ana-
lytic methods’.

The Key Characteristics of GTM

We have already discussed a detailed overview of the features of GTM in Chapter 1, but what are the key
characteristics that make it different?

Glaser and Strauss (1967) defined their method as ‘the discovery of theory from data – systematically ob-
tained and analysed in social research’. It's interesting to note that the systematic nature of the method was
emphasised from day one – and page one – of the very first book on GTM by the founders themselves.

This, of course, makes it attractive to novice researchers. As a novice researcher myself in 1997, I can re-
member stating that it offered well-signposted procedures for new researchers (Urquhart 1997). This sign-

Grounded Theory for Qualitative Research: A Practical Guide

Page 3 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

posting is clearest in Strauss (1987) and the controversial book by Strauss and Corbin (1990). Of course, this
can also be the Achilles heel of the method, in that just simply following the procedures without thought can
lead to a lack of creativity. One thing I warn my postgraduates about GTM is that it is not a failsafe recipe for
research. Even if you do follow all the prescribed steps in GTM, you cannot necessarily turn the handle and
expect a theory to drop out. Like any methodology of any kind, it requires flexibility, thought and creativity in
its application.

In Urquhart et al. (2010), we identified four key characteristics of GTM:

• the main purpose of GTM is theory building

• as a general rule, researchers should make sure that they have no preconceived theoretical ideas
before starting their research
• analysis and conceptualisation are engendered through the core process of constant comparison,
where every slice of data is compared with all existing concepts and constructs, to see if it enriches
an existing category (by adding to/enhancing its properties), forms a new one or points to a new re-
lation
• ‘slices of data’ of all kinds are selected by a process of theoretical sampling, where researchers
decide, on analytical grounds, where to sample from next.

The first characteristic implies that researchers who leverage GTM only for coding procedures are ignoring
the main purpose of the method, which is to build theory. Indeed, that is why GTM was developed in the first
place. Glaser and Strauss (1967) make a distinction between substantive theories (pertaining to the phenom-
ena at hand) and formal theories. This distinction is discussed in more detail later in this book. In developing
either type of theory, researchers need to be capable of theoretical sensitivity. Such sensitivity is based on
being steeped in the field of investigation and associated general ideas (Glaser 1978) so that researchers
understand the context in which the theory is developed. This concept of theoretical sensitivity is key. How
can we build theories ourselves unless we understand what a theory is?

The second characteristic (of having no preconceived theoretical ideas) is often held (erroneously) to imply
that researchers should not look at the existing literature before doing the empirical research. According to
Glaser (1992), the dictum in grounded theory is that there is no need to review the literature in the substantive
area under study and this idea is:

brought about by the concern that literature might contaminate, stifle or otherwise impede the re-
searcher's effort to generate categories …

Grounded Theory for Qualitative Research: A Practical Guide

Page 4 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

(Glaser 1992)

He hastens to add, though, that this applies only at the beginning. Once the theory has been sufficiently de-
veloped, researchers then need to review the literature in the substantive field and relate that literature to their
own work.

From my experience of working with postgraduates, coding for the first time, it's very hard for them to not im-
pose what they have read on the data in front of them. Being faced with the task of looking for emergent con-
cepts in the data without any help from anything other than your own mind is a scary process. Small wonder,
then, when looking for patterns in the data, people might want to fall back on what they have read already.

If, however, we privilege other theories rather than looking at the data, we lose what is for me the key delight –
and the key edge of the method: what Glaser (1992) calls ‘emergence’. The idea of emergence, for me, is that
we stay true to our data, we look for what the data is telling us. Of course, the idea of whether or not some in-
herent truth resides in the data depends on your point of view. I prefer to think of constructing meanings about
the data, but the idea that you give the data due consideration, due respect, before imposing other theories
on it makes perfect sense. It makes even more sense when we are dealing with new phenomena, such as
information technology, that have permeated most aspects of social life. For instance, if we base our under-
standing of how people interact with information technology on psychological theories and those theories are
based on large samples of American undergraduate students, how relevant is the theory we are imposing?
Far better to allow the data to tell its own story in the first instance, build a theory, then, subsequently, engage
your theory with the theory that you thought you might impose initially. You can then see if your emergent the-
ory confirms or challenges existing theories. So, potentially, GTM has a huge role to play in theory building,
in all disciplines.

The third characteristic, constant comparison, is also a key component of grounded theory. Comparative
analysis was a standard method in social research long before 1967, but in GTM it is a key part of the method.
As discussed in Chapter 1, it is the process of constantly comparing instances of data labelled in one catego-
ry with other instances of data labelled for that category.

It is an incredibly simple, but deceptively powerful, rule of thumb for analysing data. The process of constant
comparison, in my view, allows the meaning and construction of concepts to remain under review. Conscious-
ly comparing the instances of each concept allows for a fuller and more nuanced understanding of what that
category might consist of. It also, I think, allows the formation of the category to be more provisional. It is

Grounded Theory for Qualitative Research: A Practical Guide

Page 5 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

only when it is fully filled in, as it were, by many instances that we can say exactly what the meaning of that
category is. It is also helpful to use constant comparison if there is overlapping data collection and analysis
because then the category can be densified using theoretical sampling (where the emerging analysis directs
more sampling of data), so, if you like, the emerging theoretical storyline is followed.

The fourth characteristic is the selection of slices of data. This phrase was coined by Glaser and Strauss
(1967) to reflect the fact that different kinds of data give researchers different views from which to understand
a category or develop its properties. This is a liberating and interesting idea that is not always understood,
but simply means we should be constantly sampling slices of data from the phenomena in order to build the
theory out and upwards. The more diverse those slices are, the better. So, one slice of data could be field
interviews, another could be surveys. Although the 1967 book did focus on qualitative data, quantitative data
was also seen as a legitimate slice of data. Glaser and Strauss (1967) even go so far as to suggest that a
slice of data could be anecdotal or, for instance, using a national meat consumption report for a particular
professional group. When the theory is more fully formed, a conceivable slice of data might even be another
theory, as long as there is an awareness of the dangers of forcing a category down a particular road, lest that
very precious quality of grounded theory – emergence – be compromised.

These are useful characteristics to bear in mind when using GTM because they sum up what is unique about
it and give some useful guidance as to how you might use it in the field. All that said, some users of GTM
will not recognise the above characteristics because they come from the original 1967 book rather than later
versions.

We come now to the history of GTM. This history is fascinating, which is why I include it here. It's also a
contested history, which is another reason for discussing it. Researchers using GTM have to be aware of the
competing versions of grounded theory and, in particular, the well-known split between Glaser and Strauss in
1990. Each strand of GTM has its adherents, which also makes it difficult for first-time users of GTM, who may
not realise what contested territory some of these concepts occcupy. This is especially relevant for postgrad-
uate students who need to position their research in relation to a particular research community. It's important
to understand the contested principles and decide what your own position on them might be, as defending
the scholarly integrity of our work is something we should all do.

The two strands of GTM

As mentioned, any user of GTM needs to be aware of the fact that it has evolved into two distinct strands

Grounded Theory for Qualitative Research: A Practical Guide

Page 6 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

– the Glaserian and Straussian – as a result of a cataclysmic dispute between its co-founders in 1990. This
dispute went to the core of what GTM actually comprises and what it does, so understanding the dispute is
also key to understanding what GTM actually is.

The dispute was triggered by the publication of Strauss and Corbin's (1990) book. It was written in response
to their students’ requests for a ‘how to’ manual of grounded theory and contains clear guidelines and proce-
dures. It was at this point, perhaps, that the founders of grounded theory realised their views of what ground-
ed theory actually is might be different.

students of Glaser and Strauss in the 1960s and 1970s knew that the two had quite different modus
operandi, but Glaser only found out when Strauss and Corbin's Basics of Qualitative Research came
out in 1990.

(Stern 1994: 212, as quoted by Melia 1996)

Glaser reacted badly to the book and requested that it be pulled from publication. When it was not, he wrote
a correctional rejoinder, Basics of Grounded Theory Analysis: Emergence vs. forcing (Glaser 1992). For him,
the issue was nothing less than the heart and soul of GTM. He felt that Strauss and Corbin's 1990 book was
far too restrictive in the way it presented GTM. He felt strongly that to follow the procedures outlined in the
book would strangle any emergent conceptualisations and, instead, force the concepts into a preconceived
mould. He summed up his critique as follows:

If you torture the data long enough, it will give up! … [In Strauss and Corbin's method] the data is not
allowed to speak for itself as in grounded theory, and to be heard from, infrequently it has to scream.
Forcing by preconception constantly derails it from relevance.

(Glaser 1992:123)

Glaser disagreed on two fundamental issues.

• Strauss and Corbin (1990) suggested breaking down the coding process into four prescriptive steps
– open, axial, selective and ‘coding for process’ – whereas Glaser uses just three – open, selective
and theoretical coding, at incremental levels of abstraction.
• Glaser objected to the use of a coding paradigm and the ‘conditional matrix’ that are designed to
provide ready-made tools to assist with the conceptualisation process. Glaser felt that to ‘force’ cod-
ing through one paradigm and/or down one conditional path ignored the emergent nature of GTM

Grounded Theory for Qualitative Research: A Practical Guide

Page 7 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

(Glaser 1992). This makes sense, given that Glaser had suggested 18 coding families (or coding
paradigms) in his 1978 book.

So, this is a disagreement that does cut to the heart of grounded theory. The 1990 book represented a sub-
stantial departure from what had gone before, in its insistence that only one coding paradigm be used. In fact,
the 1990 book says, ‘Unless you make use of this model, your grounded theory analysis will lack density and
precision’ (Strauss and Corbin 1990). The mandatory nature of the paradigm, in retrospect, was quite extra-
ordinary. To claim that there could be only one way of relating categories and this was essential seems, in the
cold light of day, to have been an aberration in the history of GTM and one that was not necessarily heeded
by researchers.

Interestingly, this advice was modified as early as 1998:

In actuality, the paradigm is nothing more than a perspective taken toward data, another analytic
stance that helps to systematically gather and order data in such a way that structure and process
are integrated.

(Strauss and Corbin, 1998: 128)

In my own discipline, that of information systems, a colleague and I (Seidel and Urquhart 201b1) could identify
only seven papers out of 96 that applied the paradigm. Of the small percentage of papers using the paradigm,
we could see that causal relationships between categories were more frequently identified and a substantive
theory was more likely to be built than in the papers not using the paradigm.

I can remember vividly a postgraduate student of mine who rushed into my office, waving a copy of Strauss
and Corbin's 1990 book, saying, ‘This is wonderful! I want to try it!’ I did relate my own experiences of trying
to use the procedures (Urquhart 1997), when I ended up in despair as I tried to fit the paradigm to what I
was doing. We agreed that the student should try it – why not? Well, two weeks later, there was a knock at
the door and my student said, ‘It doesn't work!’ One comment I made at the time was that if the coding para-
digm recommended by Strauss and Corbin (1990) for connecting categories is a good match for the research
phenomena, it might work. Certainly there are some good examples out there of the use of this particular
paradigm (see, for instance, Galal, (2001), but my own feeling is that this only occurs when there is a good
match between the paradigm and the phenomena being investigated. Why not take advantage of the flexibil-
ity of selecting many different options for relating categories, as provided in the Glaserian version? That way,

Grounded Theory for Qualitative Research: A Practical Guide

Page 8 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

surely, you get a better match between your research problem and ways of theorising about it. For me, the
Strauss and Corbin paradigm represented, and still represents, a narrow way of thinking about what is being
investigated.

In Corbin and Strauss (2008, the updated edition of Strauss and Corbin 1998) the role of the paradigm is
further weakened in favour of emphasising a broader set of tools, named context, process and theoretical
integration. The paradigm is presented as only one of a number of ‘analytic strategies’ or ‘tools’. They write
that:

One tool for helping the researcher to identify contextual factors and then to link them with process
is what we call the paradigm. The paradigm is a perspective, a set of questions that can be applied
to data to help the analyst draw out the contextual factors and identify relationships between context
and process.

(Corbin and Strauss 2008: 89)

Throughout the book, the authors are careful to highlight that researchers must choose from a variety of an-
alytical tools and ‘make use of procedures in ways that best suit him or her’ (Corbin and Strauss, 2008: x,
Preface).

It is important, then, to read more than just the 1990 book about GTM, even though it is still widely used
and read, possibly because Glaser has self-published his books since 1978. Only when I read the original
1967 book (Glaser and Strauss 1967) and two of the most important of Glaser's subsequent books (Glaser
1978; Glaser 1992) did I realise how complex the intellectual tradition of GTM is. If only the 1990 book is
used, novice users run the danger of encountering peer reviewers for whom the type of GTM being used –
Glaserian or Straussian – really matters. It matters because understanding the dispute helps us to understand
the core principles of grounded theory. Table 2.1 lists what I feel are the key books on GTM, authored by its
founders, so that you won't fall into this trap.

Ultimately, which version of GTM is used – Glaserian or Straussian – depends on individual researchers and
their own preferences. Certainly, my view is that the Glaserian strand offers more flexibility and is closer to
the original formulation of grounded theory as put forward in the 1967 book.

The two books that are worth turning to first are Glaser (1978) and Strauss (1987). These will give you a good
sense of the two strands of GTM. Then, once you feel you have a good understanding of grounded theory,

Grounded Theory for Qualitative Research: A Practical Guide

Page 9 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

it is worth engaging with the seminal 1967 book, despite the fact that some people find it difficult to read. It
is the definitive text that started it all and gives a good sense of the original intent and form of GTM. It also
helps by elucidating the foundations from which GTM has evolved since 1967. From there, you can explore
the later works of Glaser, such as his 2005 book, which is a personal favourite of mine, simply because of the
joy and energy with which Glaser explores theory building and theoretical codes.

Let's now look more closely at the different stages of coding in the two versions.

Evolving coding procedures in GTM

Table 2.2 lists a number of different coding procedures and the books about GTM in which they appear, start-
ing with the foundational 1967 text. I'll now discuss them in turn, in the hope that this will give you enough
information to be able to choose the coding procedure that will work for you.

Table 2.1 Seminal books on GTM by its founders

Book Description

Glaser, B.G. and

Essential reading, despite it not being the most accessible of books. Melia (1996) talks of its ‘near mystical pas-
Strauss, A.L. (1967) The
sages’. As it was the first book on grounded theory, the principles on which grounded theory is based are obvious
Discovery of Grounded
and the book is notable for its concern about building theory as opposed to the mechanics of coding. It is also very
Theory: Strategies for
helpful to see the idea of grounded theory in its historical and disciplinary context.
qualitative research

Glaser, B.G. (1978) The first book in the grounded theory canon that gives a lot more detail on how the process of coding might proceed.

Theoretical Sensitivity: Introduces the idea of theoretical sensitivity – an important idea, which is that of being aware of other theories and

Advances in the how they are built. This book also discusses spacing, sampling, coding, memos, sorting and writing and provides a

methodology of ground- very important discussion on basic social processes. It introduces 18 ‘coding families’ to assist with theoretical cod-

ed theory ing.

Strauss, A. (1987) Qual- Provides advice for first-time users, especially around relating efforts to the technical literature and the process of

itative Analysis for So- coding in a group. This book also marks the first divergence into the two strands of grounded theory as only one

cial Scientists coding paradigm is proposed in this book.

Strauss, A. and Corbin

J. (1990) Basics of Probably the most widely read book on GTM, but also the most controversial. Gives very clear procedures, but, at

Qualitative Research: the same time, offers a narrower view of the method.

Grounded theory proce-

Grounded Theory for Qualitative Research: A Practical Guide

Page 10 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

dures and techniques

Glaser, B.G. (1992) Ba-

sics of Grounded Theo- This is Glaser's response to Strauss and Corbin (1990). It helps us to understand the divergent views held by Glaser

ry Analysis: Emergence and by Strauss and Corbin. It discusses in detail the significance of the issue of ‘forcing’ in GTM.

vs. forcing

Glaser, B.G. (1998) Do-

Covers practical areas of doing GTM research, such as motivation, reading the literature, forcing, generating con-
ing Grounded Theory:
cepts and more on theoretical sampling, theoretical coding, memoing, sorting and writing.
Issues and discussions

Glaser, B.G (2005) The

Grounded Theory Per- This book breaks new ground in thinking about theoretical coding and the process of relating categories. It intro-

spective III: Theoretical duces 23 new ‘coding families’ to complement the original 18 coding families in the 1978 book.

coding

Comparing incidents applicable to each category

The first important thing to note in Table 2.2 is that the original book did not set out procedures as such. It
focused more on the process than actual stages and introduced a number of important ideas in the course
of explaining it. We can see ‘comparing incidents applicable to each category’ as introducing one of the most
important ideas in grounded theory: constant comparison. Open coding – that is, the initial allocation of ideas
to data chunks – is only implied.

Table 2.2 Different GTM coding procedures

Book Suggested coding procedure

Glaser

and Comparing incidents applicable to each category (includes open coding), integrating categories and their properties (selective cod-

Strauss ing and theoretical coding), delimiting the theory (selective coding and theoretical coding), writing the theory.

1967

Glaser
Open coding, selective coding, theoretical coding.
1978

Strauss Open coding, axial coding, selective coding.

Grounded Theory for Qualitative Research: A Practical Guide

Page 11 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

1987

Strauss

and
Open coding, axial coding, selective coding.
Corbin

1990

Glaser
Open coding, selective coding, theoretical coding.
1992

Strauss

and
Open coding, axial coding, selective coding.
Corbin

1998

Charmaz
Initial coding, focused coding, axial coding, theoretical coding.
2006

Corbin
Open coding, axial coding and theoretical coding as distinct stages no longer appear, though open coding and axial coding appear
and
as terms in one chapter. The emphasis is on a broader set of tools named context, process and theoretical integration. Two coding
Strauss
paradigms are used as a foundation for context.
2008

It is the idea of constant comparison, so deceptively simple, that gives grounded theorists the edge, in my
opinion. Constant comparison has been described as a fundamental rule of thumb in GTM. To make these
comparisons between the data you are coding right now and the data you have just coded and ask, ‘To what
category does this incident or property relate?’ is the guideline that prevents inconsistent coding (when one
data chunk is coded as one thing and another very similar data chunk is coded as something else). While
the advent of data analysis software helps us compare the coding and manage the data, I believe that this
guideline is as important as ever. It ensures that researchers make their allocation of concepts to data explicit
to themselves and, more importantly, the allocations are compared to the data as a whole. Thus, we can see
hermeneutic principles of analysis also being applied to the data.

Open Coding
Open coding was first explicitly mentioned in Glaser's (1978) book. It surprises me that many people don't
refer to this book as, in many ways, it is the most important of his writings because it elaborates brilliantly on

Grounded Theory for Qualitative Research: A Practical Guide

Page 12 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

many of the key ideas in the 1967 book.

Open coding is described by Glaser (1978: 56) as ‘coding the data everyway [sic] possible’. It is the first step
in coding and deliberately ‘open’ so as not to close down any directions a future theory might take. As such it
is, along with constant comparison, a foundational technique of GTM.

The act of open coding is about attaching initial labels to your data. These are subsequently grouped into
larger codes, as the aim is to build a theory based on them. What the open codes do is flesh out what is im-
portant and point to directions in the analysis that you may not have thought of, directions suggested by the
data.

Glaser (1978) recommends coding line by line, as does Strauss (1987), for very good reasons. Charmaz
(2006) is also uncompromising in her advice on this. I can only say that the discipline of coding line by line –
that detailed consideration of the text in front of us – helps free us of our preconceptions. For example, when
I first applied grounded theory to my own PhD work into analysts and clients, I discovered that the analysts
very often attempted to frame the problem straight away. This was a unique insight for my discipline, as was
the conceptualisation of the strategies and props they used to help their clients. I firmly believe that my work
would not have been as original as it was had I used any other method than grounded theory for the analysis.

Line-by-line coding also forces a real intimacy with your data – and means that your findings are easy to de-
fend because you really know your data. That said, Glaser (1992) does say coding need not necessarily be
line by line – there are circumstances where it may not be appropriate. I do agree, because all datasets are
not created equal – some whole paragraphs may not be relevant or the data is secondary rather than primary
data. Even so, the benefits of looking at the data line by line, in my view, cannot be overstated.

Selective Coding
Selective coding is where the two strands of GTM – Glaserian and Straussian – sharply diverge. The Strauss-
ian version requires axial coding (explained in the next section) to be done first. In my view, the Glaserian
version is simpler.

Glaser (Glaser 1978) defines selective coding as the stage when coding is limited to only those categories
that relate to the core category. Future theoretical sampling also is directed by that core category. In my ex-
perience, the point at which selective coding occurs is fairly obvious, as there are no new open codes sug-
gesting themselves and definite themes are emerging. Categories become ‘saturated’ at this stage – that is,

Grounded Theory for Qualitative Research: A Practical Guide

Page 13 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

there are plenty of instances.

One point not often discussed is the fact that, at this stage, coders often end up with many multiple categories.
This is not surprising, when you consider the detailed level – line by line – that open coding starts from. So,
some grouping of categories can take place at this stage and this helps develop the abstraction of the theory.

Charmaz (2006) also makes the valuable point that the selective coding stage (which she calls ‘focused’ cod-
ing) may often prompt a return to open coding, as some interesting avenues are almost bound to occur when
we consider and group the themes emerging in the data.

Axial Coding
Strauss (1987) and Strauss and Corbin (1990) recommend a stage of axial coding following open coding. The
way I tend to think about axial coding is that it combines selective coding with the use of a coding paradigm.
The single coding paradigm recommended by Strauss and Corbin (1990) is causal conditions, context, inter-
vening conditions, action/interaction strategies and consequences.

Strauss says that, first, the codes need to be dimensionalised by laying out the properties. We can see this as
being similar to the process of selective coding, when we figure out what might be the important categories
and which of our open codes need to be elevated or otherwise combined to form those categories, as well as
which of these codes might be properties of other categories.

Strauss (1987) then says that the second element of axial coding is to hypothesise about conditions, contexts,
interactions, strategies and consequences. We can see this as a process of relating the categories theoreti-
cally, or, theoretical coding, which is explained in the next section.

My feeling is that it is hard to simultaneously figure out both the properties of a category and then how it
might relate to other categories. This is possibly why students in particular come unstuck at this point. I would
not dismiss it as a viable coding option, however, especially if researchers consider it in two stages, to sim-
plify the process. There are some examples of the successful use of axial coding in the paradigm, such as
Galal (2001), and, if you are interested, I suggest you search out this and other examples cited in Seidel and
Urquhart (2011).

There is an extensive description of axial coding in Strauss (1987), which, in my opinion, is better than that
contained in Strauss and Corbin (1998). They define axial coding as the act of:

Grounded Theory for Qualitative Research: A Practical Guide

Page 14 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

relating categories to subcategories along the lines of their properties and dimensions.

(Strauss and Corbin, 1998: 124)

Again, this is a clear indication that subcategories are involved, and its resemblance to Glaser's selective
coding. The coding paradigm is further elaborated on and put forward as conditions (causal, intervening and
contextual), actions/interactions (strategies are now put under this heading) and consequences (immediate,
cumulative, reversible, foreseen and unseen).

What is interesting is that, in Corbin and Strauss’ 2008 book, the coding paradigm loses its prominence.
Corbin says, ‘the paradigm is only a tool and not a set of directives’ (Corbin and Strauss 2008) This represents
a considerable departure from the previous books and an important one, given the paradigm's role in the split
between Glaser and Strauss. That said, the book places a great deal of emphasis on the ‘conditional/conse-
quential matrix’, to consider larger issues of context and macro conditions. In both cases, Corbin suggests
that their use should be confined to novice as opposed to experienced researchers.

Table 2.3 The evolving nature of the Strauss and Corbin paradigm

Coding paradigm Comment on evolving use of paradigm

Conditions, interactions amongst the actors, strategies In the 1987 book, it is clear that the coding paradigm is not an optional part of coding.

and consequences (Strauss 1987) Researchers are told to ‘follow the coding paradigm’ (1987: 81).

Causal conditions, context, intervening conditions, ac-

In the 1990 book, the paradigm is modified to include different types of conditions,
tion/interaction, strategies and consequences (Strauss
plus actions.
and Corbin 1990)

Conditions (causal, intervening, and contextual), ac-

tions/interactions (strategic or routine tactics), conse- In the 1998 book, conditions are clustered together, strategies are clustered under ac-

quences (immediate, cumulative, reversible, foreseen tions and consequences elaborated on.

or unseen) (Strauss and Corbin 1998)

In the 2008 book, the paradigm loses its prominence and is presented as an optional

Conditions, interactions and emotions, consequences analytic tool for novice researchers. That said, the conditional/consequence matrix,

(Corbin and Strauss 2008) used in previous editions to think about relationships between micro and macro condi-

tions, now has a more central place in coding.

Grounded Theory for Qualitative Research: A Practical Guide

Page 15 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

Theoretical Coding
Glaser (1978) describes theoretical coding as how substantive codes (the codes generated thus far pertaining
to the area under investigation) are then related to each other. This makes complete sense if we understand
that theories are constructs and relationships.

So, theoretical coding is when we relate the codes to each other and look at the nature of the relationships
between those codes. This is what builds the theory. In my own research work, I found that theoretical memos
were invaluable for helping the theorising process. A surprising (or perhaps not surprising) number of these
theoretical memos ended up being cut and pasted into my PhD thesis as, of course, I was explaining the
evolving theory to myself.

Glaser's coding paradigms give ample food for thought as to how the categories might relate – some exam-
ples are given in Table 2.4.

Of course, you can – and, I think, should be able to – generate your own coding paradigms. Grounded theory,
in my view, is, above all, about being faithful to what your analysis of the data suggests rather than shoe-
horning the data into some preconceived analytical framework, so it would go against the spirit of grounded
theory to suggest that only Glaser's coding families should be used. Given that Glaser (Glaser 2005) himself
introduced 23 more coding families, we can only assume that he, too, recognises the need to be very flexible
about how categories relate. That said, I find the coding families very useful jumping off points to think about
relationships in data and they are educational in themselves regarding theorising.

Table 2.4 A selection of Glaser's (1978; 2005) coding families

Family Comment

The 6Cs – causes, contexts, This basic coding family, together with the Strategy family, was adapted by Strauss and Corbin (1990) as

contingencies, consequences, their coding paradigm of ‘causal conditions, context, intervening conditions, action/interaction strategies and

covariances and conditions consequences’.

Process – stages, staging,

phases, phasing, progressions,

passages, gradations, transi- Glaser remarks that a process should have at least two stages. This family is similar to Spradley's ‘a stage

tions, steps, ranks, careers, or- of’.

dering, trajectories, chains, se-

quencings, etc.

Grounded Theory for Qualitative Research: A Practical Guide

Page 16 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

Dimension family – dimensions,

As Glaser says, the more we learn about a category, the more we see of its dimensions. Of all theoretical
elements, division, piece of,
codes, this is one that all researchers are likely to use. It is, of course, very similar to Spradley's ‘is a part of’.
properties of, facet, slice, sector,
It's also important to realise that, when theorising, we can privilege one dimension over another – it can be-
portion, segment, part, aspect,
come a full-blown category.
section

Glaser says, while dimensions divide up the whole, types show variations in the whole. So, for instance, you
Type family – type, form, kinds,
might have a number of styles of introducing a problem in the conversation between a systems analyst and
styles, classes, genre
client in the example in Table 3.2.

Strategy family – strategies, tac-

tics, mechanisms, managed,

way, manipulation, manoeu-
As previously remarked, the Strauss and Corbin coding paradigm seems to be a mixture of this family and
vrings, dealing with, handling,
the first family.
techniques, ploys, means,

goals, arrangements, dominat-

ing, positioning

Moment capture, when a quick

intervention is critical to causing

This is a new theoretical code introduced in Glaser's 2005 book.
an optimal outcome, such as

closing a deal

Frames, which are excavated

Also in the 2005 book. I used this idea in my own dissertation work in 1999. Glaser (2005) gives some won-
through discourse patterns and
derful nuances of causation in this theoretical code family. ‘Bias random walk’ – all variables are in a flux,
are sociocultural in nature
‘then on the introduction of a crucial variable … then of a sudden all of the variables fall into organisation’.
Causal family – a relative of the
‘Amplifying causal looping’ – ‘consequences become causes, and one sees either worsening or improving
6Cs family, it includes several
progressions or escalating severity’. ‘Conjectural causation’ – it is not always easy to identify decisive causal
aspects: 1) bias random walk, 2)
combinations. ‘Repetitive causal reproductions’ – a repeated action keeps producing the same conse-
amplifying causal looping, 3)
quences. ‘Equifinality’ – no matter what the causes and paths, the same consequence will occur. ‘Reciprocal
conjectural causation, 4) repeti-
causation’ – there is a similar interaction of effects or amplified causal looping. ‘Triggers’ – sudden causes
tive causal reproductions, 5)
that set off a consequence or set of consequences. ‘Causal paths’ – used to intervene in changing or stop-
equifinality, 6) reciprocal causa-
ping a consequence. ‘Perpetual causal looping’ – a mathematical model, an ordered calculated growth of in-
tion, 7) triggers, 8) causal paths,
creased size based on a set temporal path.
9) perpetual causal looping

It's very important to understand what theories are and how they work, so you recognise that you are building
a theory rather than just describing some interesting data analysis. Charmaz (2006) puts it well when she
cautions us to not use theoretical codes to impose frameworks on the data and be aware of a possible aura of
objectivity around those theoretical codes, as scholars would almost certainly disagree about which of those

Grounded Theory for Qualitative Research: A Practical Guide

Page 17 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

to apply. It's important to be reflective when doing theoretical coding.

Myths about GTM

Sometimes first-time users of GTM encounter certain myths about its use. There are many reasons for this,
but we've already seen in the previous sections that it is a somewhat contested intellectual tradition with dif-
ferent interpretations. Hirschheim and Newman (1991), when talking about information systems development,
use the following definition of myth:

A dramatic narrative of imagined events, usually used to explain origins or transformations of some-
thing. Also, an unquestioned belief about the practical benefits of certain techniques and behaviors
which is not supported by the demonstrated facts.

(Trice and Beyer 1984: 655 in Hirschheim and Newman 1991: 34)

Roland Barthes (1972), in his book Mythologies, argued that (cultural) myths develop not as a result of lies or
distortion but from a deceptive simplicity. So the myths I discuss here may indeed have a kernel of truth, but
their very simplicity hides a more complex truth.

One sunny day in 2005, I was sitting with a colleague, Walter Fernández, in a Brisbane café. As GTM en-
thusiasts who have both undertaken successful postgraduate projects using it, we were bemoaning the fact
that some of our colleagues clung most tenaciously to what Walter said were ‘myths’ of grounded theory, and
discouraged postgraduate students from using it as a result. This phenomenon – of departments favouring
particular methods of research and discouraging others – is, of course, not confined to grounded theory. GTM
has spread far and wide from its home discipline of sociology and not all other disciplines are equally welcom-
ing of qualitative methods. There is also the fact that GTM is still not widely used in some disciplines, so there
is sometimes a lack of knowledge and an unwillingness to supervise a student who would like to use GTM.
So, Walter and I wrote a paper (Urquhart and Fernandez 2006) to help our postgraduates defend their use of
GTM in an informed way. Below is a summary of the myths that we identified so that if you encounter them,
you too can find ways to defend your particular use of GTM.

Myth 1: Researcher as Blank Slate

If there is one issue that most academics are aware of about GTM it is the idea that it is a controversial
method, because it ignores the literature and is, thus, unacademic in some way. The idea that the GTM re-
Grounded Theory for Qualitative Research: A Practical Guide
Page 18 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

searcher is a ‘blank slate’ who launches into data collection without first looking at the literature is a particu-
larly pervasive misconception (McCallin 2003; Andrew 2006). This despite the fact that, in a footnote in the
original book, Glaser and Strauss (1967: 3) state that researchers do not approach reality as a tabula rasa
(blank slate), but must have a perspective to then abstract significant categories from the data. Dey (1993),
as mentioned earlier, speaks of the difference between an ‘open mind and an empty head’ (p.63).

According to Glaser (1992), the dictum in GTM is that there is no need to review the literature in the sub-
stantive area under study. As first suggested in Glaser and Strauss’ book of 1967 (p.37) this idea, as later
explained by Glaser ‘is brought about by the concern that literature might stifle or contaminate or otherwise
impede the researcher's effort to generate categories’ (Glaser 1992: 31) In addition, Strauss (1987) says that
the advice about delaying the scrutiny of related literature applies less to experienced researchers as they
are more practised at subjecting theoretical statements to comparative analysis.

Like most myths, the idea of researchers as a blank slate has at its base a kernel of truth. However, it is more
accurate to say that GTM research does not start with a theory to prove or disprove. It is also more helpful to
think of the literature review as being simply delayed rather than not happening at all. In fact, the grounded
theorist has an obligation, once the theory has emerged, to engage the emergent theory with the existing lit-
erature.

So, how might typical PhD students deal with the injunction to not examine the literature before coding? Such
students may actually need to review the literature for many good reasons, including passing the research
committee review. Thus, among many grounded theorists it is generally accepted that a pre-study literature
review has to be conducted to find the research problem. However, this review should be done in such a way
that the extant theories do not ‘derail the emerging theory’ (Nathaniel 2006: 40).

Martin (2006) suggests that appropriate use of the literature in GTM is a question of phasing. The first phase
is non-committal, one in which researchers develop sensitivity and find their research problem. The second
phase is integrative, in which they integrate the emergent theory with extant theories to render the new theory
in the context of existing knowledge and, thus, make the theory more valuable.

In my experience, the tactic of a preliminary (non-committal) literature review works well when using GTM. It
examines what theory exists in the area and how other people may have addressed aspects of a research
problem, but does not then impose a framework on future data collection. Importantly, this preliminary litera-

Grounded Theory for Qualitative Research: A Practical Guide

Page 19 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

ture review is conducted on the understanding it is the generated theory that will determine the relevance of
the literature. The literature review is then revisited, and extended, once the theory has been generated from
the data.

Myth 2: GTM is Inflexible

The second myth follows on somewhat from the first. The dictum about the literature is erroneously seen as
an inflexibility associated with GTM that makes it difficult to use. For instance, Allan (2003) talks of both the
difficulty of putting aside preconceptions and coding at a micro level. Allan then concludes, however, that both
these difficulties were satisfactorily resolved.

As GTM has developed, so have the procedures. Strauss (1987) and Strauss and Corbin (1990) do provide
detailed guidelines for coding data. Some of these notions of inflexibility may have come from applying
their guidelines (Strauss and Corbin 1990; 1998). For instance, Hansen and Kautz (2005), Melia (1996) and
Kendall (1999) all report difficulties in using the Strauss and Corbin paradigm.

The notion that GTM is inflexible is not borne out when one considers its widespread use in many disciplines.
It is also important to note that the procedures are commonly leveraged for the purposes of coding and build-
ing concepts as opposed to full-blown theory-building efforts. Disciplines such as health have reported that
many researchers adopt GTM for a purpose other than developing theory – generally, data analysis (Benoliel
1996). A common use for GTM in the health field, for instance, is the generation of questionnaire constructs.

Myth 3: GTM Produces Low-Level Theories that Don't do Much

Like any myth, this has in it a kernel of truth. Because GTM is grounded in the data and the coding process
starts at the word or sentence level, the theory produced tends to be rich and detailed. Also, Layder (1993),
a prominent sociologist, has repeatedly said that GTM needs to break away from focusing on micro-phenom-
ena as this prevents researchers from enriching the research with macro structures. This has tended to rein-
force the myth that GTM is about low-level theories. Thus, the general impression of GTM is that it produces
low-level theories that are difficult to ‘scale up’ (Urquhart 2001).

However, just because a low-level theory is produced does not mean that theory cannot be scaled up and,
indeed, GTM places an obligation on researchers to do so. In fact, Glaser and Strauss (1967) acknowledged

Grounded Theory for Qualitative Research: A Practical Guide

Page 20 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

from the beginning that substantive theory development can and should shade into formal theories and devot-
ed a whole chapter in their original book to this issue. They never saw GTM as only about micro theory and
both worked at organisational levels. Strauss’ interest in social arenas and social worlds led him beyond the
micro level to the ‘meso’ level (Charmaz 2006) – he talks about the obligation, having produced a substantive
theory, to wrestle with other theories (Strauss 1987). Glaser (1978), too, suggests several routes to extend-
ing and scaling up the theory, including considering similar theories and data in similar substantive areas and
how the substantive theory relates to formal models and processes.

In practical terms, it is useful to ‘scale up’ substantive theories by considering whether or not the core cate-
gories that are generated can be grouped into further concepts or themes. The important point here is that
generating a theory using GTM does not exclude researchers from the obligation of engaging their theories
with the current theories in the field and doing so is, in fact, an important element of the method.

Myth 4: GTM is Positivist/Interpretivist

In 2002, I had a most enjoyable debate with Antony Bryant about whether or not GTM carries with it philo-
sophical baggage in the shape of interpretivism or positivism inherent in the method (Bryant 2002; Urquhart
2002). The concern here is that if a myth exists about the inherent philosophical position of GTM, it may pre-
vent its use if researchers happen to come from the (supposedly) opposing paradigm. This myth is not sup-
ported by the demonstrated facts.

A qualitative method, depending on its underlying epistemology, can be seen as positivist, interpretivist or crit-
ical (Klein and Myers 1999). Therefore, GTM ‘in use’ can be influenced by different underlying epistemologies.
The fathers of GTM made no claim about the correct epistemology. Thus, we suggest in our paper (Urquhart
and Fernandez 2006) that GTM as a research method is orthogonal not only to the type of data used but also
can be appropriated by researchers with different assumptions about knowledge and how it can be obtained.
This property of the method allows researchers with dissimilar epistemological stances to succeed in using it.

In fact, GTM has been characterised as both positivist and interpretivist by various commentators. Annells
points to statements by Glaser (1992) about GTM focusing on ‘concepts of reality’ (1996: 14) and searching
for ‘true meaning’ (p. 55) as evidence of a critical realist position and inherently positivist. Madill et al. (2000)
argue convincingly that the philosophical position adopted when using grounded theory depends on the extent

Grounded Theory for Qualitative Research: A Practical Guide

Page 21 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

to which the findings are considered to be discovered within the data or the result of construction of inter-
subjective meanings. They locate the former view as Glaser's (1992) position and the latter as Strauss and
Corbin's (1998).

Charmaz's (2006) view is that GTM, in many ways, is neutral and can be seen as a container into which any
content can be poured. This would seem to be the most helpful position – simply to concentrate on GTM's
undoubted strengths for coding and theory building rather than seek an inherent philosophical bias that may
or may not be present in the method.

Myth 5: GTM takes a long time

This myth is perhaps less pervasive, but new PhD students will still come across it. It has its root in the idea
that qualitative research takes longer than quantitative research because of the nature of data collected. I
have even come across PhD students who have been advised to do a statistical PhD because it is quicker
and save their qualitative yearnings for later, once they have their PhD!

So, is there any truth in the myth that a grounded theory PhD might take a long time to do? Certainly Walter
Fernández and I have had students who have done grounded theory PhDs and it seems to have taken no
longer than other PhDs. In fact, one advantage of doing a grounded theory PhD, we have observed, is that
there is a wealth of material in it that can be used in subsequent publications – it provides a rich platform for
a future academic career.

Perhaps it is not a question of time, given that all the students I have supervised doing grounded theory have
finished on time. I would hazard a guess that, even if the analysis phase takes longer than other methods, the
level of engagement and associated theorising means that the write-up is quicker.

The likely explanation for the origin of this myth is that there is a perceived level of difficulty with qualitative
analysis in general. One student of mine (now an esteemed colleague) described the process as difficult, a
lot of hard work, but ultimately worthwhile. So, while grounded theory takes no longer than other methods of
analysis, it is perhaps more challenging – and, I would add, more rewarding!

Summary

Grounded Theory for Qualitative Research: A Practical Guide

Page 22 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

• This chapter discussed the discovery of grounded theory in 1967, how it all started and what the
key characteristics of GTM are.
• I then spent some time talking about the fact that GTM has evolved into two major strands – the
Glaserian and Straussian – due to a split between the founders in 1990. It is important to understand
that split – I would go so far to say, if you understand the dispute and its ramifications, you under-
stand what grounded theory is about.
• I also traced the key publications in the history of grounded theory. While not wanting to be pre-
scriptive about what people should read, it does seem to me that the grounded theory canon needs
to be engaged with properly. It has such an extensive intellectual tradition that just reading one book
by the founders is not sufficient.
• The chapter then went on to explore the different coding stages in GTM, depending on which strand
is followed – Glaserian or Straussian. Now the use of the coding paradigm that caused the split is
advisory rather than mandatory (Corbin and Strauss 2008), the Straussian strand has returned to
some flexibility, and perhaps the strands can be seen as more equivalent.
• Finally, I concluded by discussing some myths about GTM that first-time users can encounter.
These myths are not necessarily malevolent in character, but can prevent people from using what I
think is an incredibly useful method. The major barrier to GTM use from a postgraduate perspective
is the positioning of a literature review, when grounded theory suggests that this be delayed. Under-
taking a non-committal literature review can help such students square the circle in this situation.
• As GTM is a living tradition, many adaptations and further debates are both possible and inevitable.
The aim of this chapter has been to give you a good understanding of the foundations of GTM.

Exercises

1. Do a literature search in your own discipline to find out how GTM has been used to date. Can you
identify which strand is being used and which intellectual traditions are being referred to? Have there
been any debates in your discipline about the use of GTM? What do those debates tell you?
2. If you are working in a group, set up a role-play where one of you plays a sceptical senior colleague
who talks about the barriers to using GTM and the other a new researcher who wants to use GTM
for a piece of PhD research. What issues might come up? Can you think of any more myths about
GTM you might encounter?

Web Resources
Grounded Theory for Qualitative Research: A Practical Guide
Page 23 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

www.groundedtheory.com This is Dr Barney Glaser's official site. The emphasis here is on what
he calls ‘classic grounded theory’. As Glaser is still writing and working on GTM, there is a very real
sense that this website represents a living and evolving method.

Further Reading

Melia, K. M. (1996) ‘Rediscovering Glaser’, Qualitative Health Research, 6(3): 368–73. This ar-
ticle, by Kath Melia, Professor of Nursing Studies at Edinburgh University and one of the early pi-
oneers of grounded theory in nursing, is the best I have ever read about the split between the
founders.
Walker, D. and Myrick, F. (2006) ‘Grounded theory: An exploration of process and procedure’,
Qualitative Health Research, 16(4): 547–59. Discusses the differences between coding procedures
in the Glaserian and Straussian strands in detail.

Frequently asked Questions

Does ignoring the literature not carry the risk of reinventing the wheel? If we investigate an
area where there is a lot of literature, is there a danger that we will just come up with a similar
theory and find nothing new?

This is an excellent question, which I will choose to answer in several ways. First, while the GTM dictum is to
leave the literature review until later, there is, indeed, as mentioned earlier, a big difference between an open
mind and an empty head! I would suggest it is unlikely, in an established discipline, that you would not already
have some idea of prevailing theories and issues. Second, if you do choose to do a non-committal literature
review, you provide yourself with a safety net in this instance. Finally – and this is perhaps the most important
point – I have never come across an application of GTM that does not either contribute theory or extend ex-
isting theory in some way. I think this is because GTM, by dint of its systematic coding process, necessitates
a very close look at the data and, because of this, something new is invariably discovered.

Which strand of GTM is best to use – the Straussian or Glaserian?

My own preference is for the Glaserian because I see it as more flexible and closer to the original ideas
of grounded theory, as advanced in the 1967 book. That said, I have seen successful applications of the
Straussian version and it's also important to remember that Corbin and Strauss’ (2008) book is much more

Grounded Theory for Qualitative Research: A Practical Guide

Page 24 of 25
SAGE SAGE Research Methods
© Cathy Urquhart 2013

flexible about procedures. If you are a postgraduate student, you'll also have to consider which strand is more
popular in your discipline and any previous work done by your supervisor using GTM.

https://dx.doi.org/10.4135/9781526402196

Grounded Theory for Qualitative Research: A Practical Guide

Page 25 of 25