DRO
Deakin Research Online,
Deakin University’s Research Repository Deakin University CRICOS Provider Code: 00113B
Grounded theory methodology in information systems research
Citation:
Wiesche, Manuel, Jurisch, Marlen C, Yetton, Philip W and Krcmar, Helmut 2017, Grounded
theory methodology in information systems research, MIS quarterly, vol. 41, no. 3, pp. 685-
701.
DOI: http://dx.doi.org/10.25300/MISQ/2017/41.3.02
©2017, The Authors
Reproduced by Deakin University under the terms of the Creative Commons Attribution Licence
Downloaded from DRO:
http://hdl.handle.net/10536/DRO/DU:30105882
METHODS ARTICLE
GROUNDED THEORY METHODOLOGY IN
INFORMATION SYSTEMS RESEARCH
1
Manuel Wiesche
Technische Universität München, Chair for Information Systems,
Munich, GERMANY {wiesche@in.tum.de}
Marlen C. Jurisch
City of Munich, Department for IT-Strategy, IT-Controlling, & IT Governance
Munich, GERMANY {marlenjurisch@gmail.com}
Philip W. Yetton
Deaken University, Department of Information Systems and Business Analytics,
Burwood, Victoria, AUSTRALIA {philip.yetton@deakin.edu.au}
Helmut Krcmar
Technische Universität München, Chair for Information Systems,
Munich, GERMANY {krcmar@in.tum.de}
Grounded theory methodology (GTM), with its espoused goal of theory development of novel phenomena, has
found broad application in Information Systems (IS) research. To investigate how GTM is applied in IS
research and how the research contributions are contingent on those applications, we review 43 GTM-based
articles in major IS and related journals. Ten of the articles develop theory. The other 33 articles use GTM
to develop models and rich descriptions of new phenomena as their theoretical contribution. We show that
each of the three forms is valuable to the IS community. For example, studies that develop theories and models
are highly cited in the IS literature. We identify nine GTM procedures that are applied in various combinations
to develop the three forms of research contribution. Treating GTM as a portfolio of the nine procedures, we
examine the implications for the research contribution of adopting the core GTM procedures compared with
a partial portfolio of those procedures.
1
Keywords: Grounded theory methodology, explorative, qualitative, research method, procedures, portfolio
approach
1
Fred Niederman was the accepting senior editor for this paper.
The appendix for this paper is located in the “Online Supplements” section of the MIS Quarterly’s website (http://www.misq.org).
© 2017 The Authors. Published by the Management Information Systems Research Center at the University of Minnesota. This is an open access article under
the terms of the Creative Commons Attribution CC BY License, which permits use, distribution, and reproduction in any medium, provided the original work
is properly cited.
MIS Quarterly Vol. 41 No. 3, pp. 685-701/September 2017 685
Wiesche et al./Grounded Theory Methodology in IS Research
Introduction
Grounded theory methodology (GTM) is designed to enable
the discovery of inductive theory. It “allows the researcher to
develop a theoretical account of the general features of a topic
while simultaneously grounding the account in empirical
observations or data” (Martin and Turner 1986, p. 141).
Developed four decades ago, GTM has become one of the
most frequently adopted qualitative research methods in social
science research (Morse 2009).
The GTM literature provides researchers with guidelines,
advice, and perspectives regarding its use and to ensure the
rigor of the research contribution (see Charmaz 2011; Glaser
1978; Strauss and Corbin 1990; Urquhart 2002). The method
is particularly relevant for research on issues for which
limited prior research has been conducted and for which
theory building is needed (Fernandez 2004; Lehmann 2010;
Seidel and Urquhart 2013). In IS research, GTM has been
chosen frequently to study technological change and socio-
technical behavior in emerging research domains (Birks et al.
2013; Matavire and Brown 2013; Urquhart and Fernandez
2006).
However, for two reasons, we contend that IS research has not
exploited GTM to its full potential. One is that many studies
do not develop theory, which is the espoused goal of GTM
(Lehmann 2010; Urquhart et al. 2010). While each of the
three forms of GTM results makes a theoretical contribution,
we make a formal distinction between developing new theory
and developing models and rich descriptions of new phenom-
ena. The other reason is that there is ambiguity concerning
how GTM should be applied in IS research (Birks et al. 2013;
Sarker et al. 2013; Seidel and Urquhart 2013; Urquhart and
Fernandez 2006).
In practice, the various applications of GTM challenge
researchers, journal editors, and reviewers (Hughes and Jones
2003; Morse 2009). IS researchers would benefit from addi-
tional guidance on the appropriate applications of GTM
procedures, particularly with regard to the consequences of
the procedures chosen for the form of the research contribu-
tion (Sarker 2007). Editors and reviewers would benefit from
a more nuanced understanding of the different forms of
research contribution from GTM-based IS research (Sarker et
al. 2013).
To explore these issues, we investigate how GTM is applied
in IS research and how research contributions are contingent
on the procedures adopted. To do this, we present a review of
43 GTM-based IS articles published before April 2013. We
classify the research contribution of each article as taking one
of three forms: theories, models, and rich descriptions of
phenomena.
In making the distinction among the three forms of research
contribution, we define theories to include the definitions of
the relevant variables, the relationships among those vari-
ables, the justifications for those relationships, and the bound-
aries of the theory (Sutton and Staw 1995; Whetten 1989).
Models include the definitions of the relevant variables and
the relationships among those variables but do not fully justify
those relationships and specify their boundaries. Thus,
models are frequently the basis for theory development
(Markus and Robey 1988; Sutton and Staw 1995). Rich
descriptions are narratives based on empirical observations
without abstraction (Van Maanen 1990).
The form of the research contribution is contingent
2
on the
GTM procedures adopted. The development of theory is fre-
quently contingent on the adoption of the full range of GTM
procedures. In contrast, models and rich descriptions are
frequently reported by studies that adopt a partial portfolio of
GTM procedures. In addition, the analysis shows that GTM
articles that develop theory are highly cited compared with
non-GTM articles published in the same journal in the same
year. GTM articles that develop models are also highly cited.
Citation counts are lower for articles that develop rich
descriptions.
The rest of the paper is organized into four sections. First, a
framework is developed that includes the research contribu-
tion, the GTM procedures adopted, and the GTM context that
guides our analysis. Second, the study design is presented.
Third, we report the findings for and analysis of the GTM
procedures adopted, their effects on the research contribution,
and the influence of those contributions on the IS literature.
Fourth, we discuss the findings and the implications for IS
research.
A Framework for Investigating
GTM in IS Research
GTM articles differ in the form of the research contribution
that they make. These forms range from developing theory,
the espoused goal of GTM, to publishing rich descriptions of
new phenomena (Birks and Mills 2011; Lehmann 2010;
Urquhart et al. 2010). For this research, we classify GTM
research contributions in IS as the development of theories,
models, and rich descriptions.
2
In using the word contingent, we do not claim that the findings test for
causality and its direction in the relationship. Rather, accepting the assump-
tion in GTM that the methodology is the mechanism on which the research
contribution is contingent, we show how, in this particular relationship and
in the subsequent relationships reported in this paper, the research contribu-
tion is contingent on the GTM procedures adopted.
686 MIS Quarterly Vol. 41 No. 3/September 2017
Wiesche et al./Grounded Theory Methodology in IS Research
Table 1. Three Forms of Research Contribution
Form of Research
Contribution Description Reference
Theory Statements of descriptions, definitions of variables, their
relationships, justifications for those relationships, and the
boundaries of the theory.
Sutton and Staw 1995;
Whetten 1989
Model Definitions of abstract variables and their relationships. Markus and Robey 1988;
Sutton and Staw 1995
Rich description Narratives of empirical observations without abstraction. Hambrick 2007; Van Maanen 1990
It is generally accepted that the application of GTM requires
tailoring GTM procedures to the research context (Hughes
and Jones 2003; Morse 2009). However, disagreement exists
regarding the degree of tailoring that is appropriate (Sarker
2007). Positions range from loose (Hood 2007; Locke 1996;
Strauss 1987) to strict (Birks et al. 2013; Glaser 1992;
Goulding 1999) adaptations. However, there is limited empir-
ical research on how the various positions affect the form of
the research contribution that is developed.
In addition, the application of GTM in IS depends on other
contextual factors, including access to the research site and
the duration of the study. Here, we develop a framework to
review the application of GTM in IS research and the effects
of differences in GTM practices on the form of the published
research contribution. The framework includes three forms of
research contribution, nine GTM procedures, and six con-
textual factors.
Research Contributions of GTM Articles in IS
While the espoused goal of GTM is the development of
theory, GTM studies in IS do not exclusively develop theory.
Instead, as Table 1 shows, these studies make three research
contributions, namely, theories, models, and rich descriptions.
Sutton and Staw (1995) and Whetten (1989) define theories
3
to include definitions of abstract variables, their relationships,
justifications for those relationships, and boundaries that limit
the scope of the theory. Similarly, Gregor (2006) and Rivard
(2014) define theories as consisting of descriptions, models,
justifications, and boundary conditions. Thus, theories com-
prise both descriptions and explanations. Descriptions are
empirical narratives that illustrate the author’s arguments.
Explanations specify the logic that justifies the selection of
factors, the proposed causal relationships, and the boundaries
of the theory (Bacharach 1989; Whetten 1989).
Markus and Robey (1988) and Sutton and Staw (1995) define
models as definitions of abstract variables and their relation-
ships. These variables and relationships are based on gener-
alizations from data. Critically, models do not provide
explanations for the relationships (Davis and Marquis 2005;
Silverman 2014). In this sense, models are pre-theoretical
representations of reality. Typically, the relationships among
concepts, categories, and properties are formulated as hypoth-
eses or propositions in a particular subject area (Glaser 1998;
Whetten 1989).
Rich descriptions are narratives based on empirical observa-
tions without abstraction (Van Maanen 1990). They include
objects, people, systems, activities, and events that occur in a
particular area of study (Bacharach 1989). These narratives
highlight the patterns and clusters among attributes within the
data (Birks and Mills 2011; Hambrick 2007; Lehmann 2010).
Essentially, rich descriptions are narrative reports of the major
events, conceptualizing the emerging relationships among the
variables involved (Becker 1993). These descriptions illum-
inate a phenomenon and help advance the practical relevance
of a theory or the need to develop a theory (Sutton and Staw
1995; Van Maanen 1990).
Although Glaser and Strauss (1967) treat conceptual cate-
gories, their properties, and generalized relations as elements
of theory, they do not provide a formal definition of theory
(Charmaz 2006). Glaser’s (1992) notion of fit, work, and
modifiability captures both descriptive and explanatory
elements. A theory fits if categories and properties capture
the realities under study (description). A theory works if it
explains major variations in behavior in the area (explana-
tion). A theory is modifiable if it can include new data and
new concepts (Glaser 1978).
3
The grounded theory literature in sociology differentiates substantive and
formal theories (Glaser 1978; Kearney 2007). Substantive theories are
developed for an empirical area of inquiry, whereas formal theories are
developed for a conceptual area of inquiry, comprising the analysis of several
sets of data (Glaser 2006). In IS, however, GTM produces mainly substan-
tive theories because IS has a culture of close domain fit as an applied
discipline (Sarker 2007; Urquhart et al. 2010).
MIS Quarterly Vol. 41 No. 3/September 2017 687
Wiesche et al./Grounded Theory Methodology in IS Research
Table 2. GTM Procedures for Data Collection and Analysis
GTM
Procedure Description Reference
Theoretical
sampling
The process of selecting the data to be collected based on the analysis of
previously collected data.
Glaser and Strauss 1967
Role of prior
theory
The degree to which theories from the extant literature are used to inform
GTM data collection and analysis.
Glaser and Strauss 1967;
Strauss and Corbin 1990
Open coding The act of attaching initial labels to all available data. Glaser 1978
Axial coding The detailed analysis of one category (around the “axis” of the category). Strauss 1987
Selective
coding
Coding limited to identifying only those instances related to the core
category.
Glaser 1978
Theoretical
coding
A coding step that relates the substantive categories generated from
selective coding to one another.
Glaser 1978
Constant
comparison
The process of constantly comparing any unit of data in one category with
another unit.
Glaser and Strauss 1967
Memoing Write-ups of ideas about concepts, categories, and the relationships
among them that occur during the analysis.
Glaser 1978
Coding
paradigm/
coding families
The Straussian paradigm is a pattern of analysis for examining data
regarding conditions, interactions, tactics, and consequences. Glaser
suggests a set of broader theoretical options, i.e., coding families.
Glaser 1978, 2005; Strauss
and Corbin 1998
GTM Procedures in IS Research
There is no unique, generally accepted set of GTM procedures
to guide the coding process during data collection and
analysis. The GTM procedures presented in Table 2 identify
two procedures for data collection and seven procedures for
data analysis. Theoretical sampling and the role of prior
theory guide the data collection.
Theoretical sampling reduces sampling bias, and increases
data coverage and the saturation of categories (Glaser 1992).
The role of prior theory captures the degree to which
researchers are free of preconceived theoretical concepts and
relationships when they enter the field (Baker et al. 1992;
Sarker 2007; Suddaby 2006; Urquhart and Fernandez 2006).
However, it is acceptable to use prior theory to motivate the
relevance of a particular study, outline the research gap, and
link the GTM results to the existing body of knowledge
(Glaser 1992; Strauss and Corbin 1998; Urquhart 2002).
Although the literature contains an extensive discussion of
second-generation approaches to GTM (see Charmaz 2011;
Clarke 2005), GTM studies in IS generally follow either the
Glaserian or Straussian coding approach (Glaser 1978;
Strauss and Corbin 1998). Both begin with open coding, but
they differ in subsequent steps. Open coding is the initial
line-by-line coding of all data, which is often documented by
example codes and the total number of open codes (see
Smolander et al. 2008; Strong and Volkoff 2010).
GTM studies that follow Glaser (1978) in subsequent steps,
apply selective coding by identifying categories that are
related to the core category. Selective coding is documented
by reporting on examples and explaining the reasons for con-
ducting certain steps in abstraction (see Lee 2001). GTM
studies that follow the Straussian coding procedure (Strauss
1987) conduct axial coding, which is performed to develop a
deeper knowledge of all categories, as an additional step
before applying selective coding. The studies reporting axial
coding describe the properties of the categories and the
implications of axial coding (Smolander et al. 2008).
Theoretical coding, which is the last step in Glaserian coding,
identifies relationships between categories that are associated
with the core category (Glaser 1978). Theoretical coding is
documented by examples and the provision of additional de-
tails regarding their relationships (Gasson and Waters 2013).
Constant comparison is a procedure that guides analysis. The
researcher systematically compares any unit of data with an-
other unit to ensure that the discovery is grounded in rigorous
coding and systematic procedures (Charmaz 2006). Studies
report using constant comparison to generate concepts or cate-
gories (see Levina and Ross 2003) to generate new properties
(see Gasson and Waters 2013), and to ensure the best fit
between the analysis and the data (see Zahedi et al. 2006).
Memoing is a technique that is used to note theoretical ideas
during data analysis and communicate insights from the data
688 MIS Quarterly Vol. 41 No. 3/September 2017
