PUTTING THE THEORY BACK INTO GROUNDED THEORY

doi:10.1111/j.1365-2575.2009.00328.x Info Systems J (2010) 20, 357–381

357

Putting the ‘theory’ back into grounded theory: guidelines for grounded theory studies in information systems Cathy Urquhart,* Hans Lehmann† & Michael D. Myers‡ *Department of Information Systems and Operations Management, University of Auckland, Private Bag 92019, Auckland, New Zealand, email: [email protected], † School of Information Management, Victoria University, Wellington, New Zealand, email: [email protected], and ‡Department of Information Systems and Operations Management, University of Auckland, Private Bag 92019, Auckland, New Zealand, email: [email protected]

Abstract. Over the past decade, there has been increasing interest in the use of grounded theory in information systems research. Grounded theory is a qualitative research method that seeks to develop theory that is grounded in data systematically gathered and analysed. The purpose of this paper is to suggest guidelines for grounded theory studies in information systems. Our guidelines are based on a framework for theorizing in grounded theory studies that focuses on conceptualization and theory scope. Our hope is that the guidelines will help to raise the quality and aspirations of grounded theory studies in information systems. Keywords: grounded theory, research methods, theory building, guidelines for grounded theory

INTRODUCTION

Over the past decade, there has been increasing interest in the use of grounded theory in information systems research (Howcroft & Hughes, 1999; Hughes & Howcroft, 2000; Urquhart, 2001; 2007; Lundell & Lings, 2003; Bryant et al., 2004; Lings & Lundell, 2005). Grounded theory is a qualitative research method that seeks to develop theory that is grounded in data systematically gathered and analysed. According to Martin & Turner (1986), grounded theory is ‘an inductive, theory discovery methodology that 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’. The major difference between grounded theory and other qualitative research methods is its specific approach to theory development – grounded theory suggests that there should be a continuous interplay between data collection and analysis.

© 2009 Blackwell Publishing Ltd

358

C Urquhart et al.

In information systems, grounded theory has proved to be extremely useful in developing context-based, process-oriented descriptions and explanations of information systems phenomena (Myers, 1997; Goulielmos, 2004). It offers relatively well-signposted procedures for data analysis, and potentially allows for the emergence of original and rich findings that are closely tied to the data (Orlikowski, 1993). It is this last feature that provides researchers with a great deal of confidence, as for each concept produced, the researcher can point to dozens of instances in the data that relate to it. However, grounded theory studies in information systems have been criticized for having a relatively low level of theory development. Many grounded theory studies in information systems use grounded theory only as a coding method, and, indeed, the term ‘grounded theory’ itself has almost become a blanket term for a way of coding data (Hughes & Howcroft, 2000; Bryant et al., 2004; Urquhart, 2007). Interestingly, this particular usage of the grounded theory method is not limited to the field of information systems. Scholars in other fields have highlighted exactly the same issue, of the grounded theory method being viewed primarily as a way of coding data rather than a method for generating theory (Becker, 1983; Benoliel, 1996; Green, 1998; Elliott & Lazenbatt, 2005). We believe that this use of grounded theory – while appropriate in some cases – is somewhat limited. Grounded theory is not just a coding technique, but offers a comprehensive method of theory generation. In fact, one of the attractions of grounded theory for information systems researchers is the promise that it will help us to develop new theories of information systems phenomena – theories that are firmly grounded in empirical phenomena. Given the calls for information systems researchers to focus more on theory development (Watson, 2001; Weber, 2003), we suggest that grounded theory can be used to help generate theories in information systems. Hence, a key question that this paper seeks to address is: ‘How can the grounded theory method be leveraged to build theory in information systems?’ The purpose of this paper, therefore, is to suggest guidelines for conducting and evaluating grounded theory studies in information systems. Our hope is that the guidelines will help to raise the quality and aspirations of grounded theory studies in information systems, and, as a consequence, contribute to theory development in the field. This paper should be of interest to all information systems researchers using or considering using grounded theory, and to all other information systems researchers who, while not using grounded theory themselves, might like to understand its potential contribution. This paper is organized as follows. In the next section, we clarify the nature of grounded theory and outline its theoretical and philosophical foundations. In the third section, we discuss the issue of theory building in information systems, and what grounded theory can contribute to theory building in the IS discipline. In the fourth section, we propose a framework for theorizing using grounded theory. In the fifth section, we discuss our suggested guidelines for the conduct and evaluation of grounded theory studies based on the framework. In the sixth section, we use these guidelines to analyse three grounded theory studies in information systems. The final section is the discussion and conclusions.

© 2009 Blackwell Publishing Ltd, Information Systems Journal 20, 357–381

Guidelines for grounded theory studies in information systems

359

GROUNDED THEORY METHOD – AN OVERVIEW

Barney Glaser and Anselm Strauss published a book entitled ‘The Discovery of Grounded Theory’ in 1967 (Glaser & Strauss, 1967). This book outlined a research methodology that aimed at systematically deriving theories of human behaviour from empirical data. It was a reaction against ‘armchair’ functionalist theories in sociology (Dey, 1999; Kendall, 1999). Several more books and articles by the co-originators followed, which developed and later debated the method (Glaser, 1978; 1992; 1995; 1998; 1999; 2001; Strauss, 1987; Strauss & Corbin, 1990; 1994; 1997). Following the publication of this seminal work in 1967, grounded theory spread fairly quickly as a qualitative research method within the social sciences and many other fields. For example, Benoliel (cited in Dey, 1999, p. 412) says there was a 70-fold increase in published papers with ‘grounded theory’ as a keyword in the health field over the previous decade. By the mid-1990s, the methodological procedures of grounded theory had permeated qualitative research to such an extent that Miles and Huberman labelled it a ‘ “common feature” [of qualitative] analytic methods’ (Miles & Huberman, 1994, p. 9). There are various definitions of grounded theory. The earliest, a process definition by the creators themselves, defines it as ‘the discovery of theory from data – systematically obtained and analysed in social research’ (Glaser & Strauss, 1967, p. 1). A more detailed definition is as follows: The methodological thrust of grounded theory is toward the development of theory, without any particular commitment to specific kinds of data, lines of research, or theoretical interests . . . Rather it is a style of doing qualitative analysis that includes a number of distinct features . . . and the use of a coding paradigm to ensure conceptual development and density (Strauss, 1987). From these various definitions, we can discern four distinctive characteristics of the grounded theory method. These are as follows: 1 The main purpose of the grounded theory method is theory building. 2 As a general rule, the researcher should make sure that their prior – often expert – knowledge of the field does not lead them to preformulated hypotheses that their research then seeks to verify – or otherwise. Such preconceived theoretical ideas could hinder the emergence of ideas that should be firmly rooted in the data in the first instance. 3 Analysis and conceptualization are engendered through the core process of joint data collection and constant comparison, where every slice of data is compared with all existing concepts and constructs to see if it enriches an existing category (i.e. by adding/enhancing its properties), forms a new one or points to a new relation. 4 ‘Slices of data’ of all kinds are selected by a process of theoretical sampling, where the researcher decides on analytical grounds where to sample from next. The first characteristic implies that researchers who use grounded theory only as a way of coding data are neglecting the main purpose of the method – which is to build theory. Theory


360

C Urquhart et al.

building is why grounded theory was developed in the first place. In developing theory, the researcher needs to be capable of theoretical sensitivity. Theoretical sensitivity is based on being steeped in the field of investigation and associated general ideas, so that a researcher understands the context in which the theory is developed (Glaser, 1978). The second characteristic – of no preformulated hypotheses – underscores the first – i.e. that theory building, not theory verification, is the main and only aim of grounded theory. However, it is often held to imply that the researcher should not look at the existing literature before doing the empirical research. This injunction is mainly designed to ensure that the researcher does not impose ideas from the literature on that coding. If the researcher starts with an existing theory, then the aim of the grounded theory method is to enhance the theory, widen its scope or in other ways improve it – but not to verify or falsify it. In a footnote in the original 1967 book, Glaser and Strauss (1967, p. 3) state that the researcher does not approach reality as a tabula rasa but must have a perspective that will help him or her abstract significant categories from the data. Dey (1999) speaks of the difference between an ‘open mind and an empty head’ – both he and we believe that the founders of grounded theory inclined to the former position. The third characteristic requires joint interaction between data collection and comparison. Although comparative analysis has been a standard method in social research long before 1967, the specific rigour and the level of detail demanded by grounded theory are significantly greater. Glaser & Strauss (1967, p. 43) emphasize that the data collection, coding and analysis need to be done together because separating these operations might hinder the development of theory. They give the example of a fresh analytic idea emerging in the coding that may redefine the collection but is ignored due to pre-established rules or routine – thus stifling the generation of theory. The idea of joint interaction between data collection and analysis is central in grounded theory. As for the fourth characteristic, the term ‘slices of data’ was coined by Glaser and Strauss to reflect the fact that different kinds of data give the researcher different views from which to understand a category or to develop its properties. They say ‘these different views we have called slices of data’ (Glaser & Strauss, 1967, p. 65). Theoretical sampling is where the researcher uses the categories, concepts and constructs established so far to direct further data collection (Glaser, 1978). In the original book, a whole chapter is devoted on how to select sample groups to aid development of the emerging theory (Glaser & Strauss, 1967).

Philosophical foundations of grounded theory There is considerable disagreement and debate with regard to the underlying philosophical assumptions of grounded theory. Grounded theory belongs to the realm of qualitative empiricism and has been variously described as positivist, interpretive or critical. However, Bryant (2002) is critical of the founders for their phenomenalist approach, which assumes that a theory is just waiting to be discovered in the data. Clearly, this idea does not take into account the subjectivism of coding. We tend to see this philosophical uncertainty as simply reflective of the intellectual climate of the time in which Glaser and Strauss introduced grounded theory. Holton



361

(2007) says that Charmaz (2006) identifies the crux of the matter as a lack of clarity in the seminal work of 1967, and suggests that a search for a position is futile.1 Following Myers (1997), we take the view that grounded theory is primarily a qualitative research method for gathering and analysing data. As a research method, grounded theory is independent of the underlying epistemology. This means that grounded theory is itself, as Glaser describes, ‘paradigmatically neutral’ (Glaser, 2001). It can be used in positivist studies (Lehmann, 2003), interpretive or critical studies (Annells, 1996; Urquhart, 2001; CecezKecmanovic et al., 2008). Grounded theory’s neutrality also makes it useful for mixedparadigm research (Charmaz, 2005). A key point we wish to make here is that a researcher’s own ontological and epistemological position will impact on their coding and analysis of the data and the way in which they use grounded theory (Madill et al., 2000). Our suggested guidelines, however, apply to all kinds of grounded theory.

Two strands of grounded theory In subsequent years the grounded theory method has evolved into two distinct variants, one favoured by Glaser, the other by Strauss (Melia, 1996). A very public disagreement between these two co-founders of grounded theory occurred on the publication of Strauss and Corbin’s book in 1990 (Strauss & Corbin, 1990). This book was written in response to their students’ requests for a ‘how to’ manual of grounded theory, and contains clear guidelines and procedures. In Glaser’s view, this formalization is far too restrictive, to the extent that it may strangle any emergent conceptualizations and instead force the concepts into a preconceived mould. Glaser felt so strongly about the Strauss and Corbin book that he requested it to be pulled from publication, and when it was not, wrote a correctional rejoinder ‘Emergence vs. Forcing: Basics of Grounded Theory Analysis’ (Glaser, 1992). He sums up his critique as follows: If you torture the data long enough, it will give up! . . . [In Strauss & 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, p. 123). Glaser disagreed on two fundamental issues. First, Strauss & 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. Second, Glaser objected to the use of a coding paradigm and the ‘conditional matrix’, which are designed to provide ready-made tools for the conceptualization process. Glaser pointed out that to ‘force’ coding through one paradigm and/or down one conditional path was not grounded theory, but conceptual description, which ignored 1

For those readers interested in philosophical reinterpretations of grounded theory, we can do no better to point interested readers to Charmaz (2006) for a constructivist re-rendering of grounded theory, and to Clarke (2005) for a post-modern view.


362

C Urquhart et al.

the emergent nature of grounded theory (Glaser, 1992). Also, the coding paradigm used by Strauss and Corbin – which suggests that the researcher looks at context, conditions, action/ interactional strategies, intervening conditions and consequences for the purposes of establishing categories and relationships – can be further critiqued as a departure from traditional grounded theory. First, this coding paradigm provides only one particular view of a phenomena. By contrast, Glaser (1978) suggests 18 coding families, which cover ideas like dimensions and elements, mutual effects and reciprocity, social control, recruitment and isolation, and many other ideas for categories and relationships. Second, we have found that the insistence on a phase of axial coding, where categories and relationships are considered simultaneously using the coding paradigm, causes real difficulty for some researchers, especially novices. Selective coding, followed by theoretical coding, in our view, allows for more abstract theorizing, as has been noted by Kendall (1999). Given that this is such a well-documented disagreement between the two co-founders of grounded theory, we believe that information systems researchers need to be aware which version of grounded theory they are using. The Strauss and Corbin (1990) book is arguably the most widely known, and regarded as the most accessible. However, it describes only one version of grounded theory, and has also been described as rather formulaic and overburdened with rules (Melia, 1996; Kendall, 1999).

The generation of grounded theory The process of generating a grounded theory is summarized in Figure 1. A researcher begins a grounded theory study with ideational constructs, such as ‘hunches’ (Miles & Huberman, 1984), for investigation. It is important to note that despite the injunction to try to avoid having any preconceived theoretical ideas before starting the research, these seed concepts or early hunches ‘can come from sources other than data’ (Glaser & Strauss, 1967, p. 6). These seed concepts help a researcher to select an area of enquiry and define the topic. The area of enquiry is called the ‘substantive area’ in grounded theory terminology. Next, the researcher takes ‘slices of data’ (Glaser & Strauss, 1967) from the area of enquiry and codes them into conceptual categories. These slices of data can come from many different sources, and can be collected using many different data collection methods. This of course provides an opportunity for corroboration or triangulation of the data. As the first element of a grounded theory, these conceptual categories are first described by their properties. Using additional slices of data, the categories are further conceptualized into theoretical constructs by establishing ‘relations’ between them (Glaser & Strauss, 1967, p. 35). Constant comparison with previous data, categories, concepts and constructs is the key. Additional data are acquired using theoretical sampling until the existing categories are ‘saturated’ (i.e. there are no more instances of them in the data), and until no more new conceptual categories or relations emerge. The ‘saturated’ concepts are then reduced as much as possible to the relationships between core categories, which then form a ‘grounded’ theory. The grounded theory that is produced is thus firmly anchored in the data that led to its formulation.



'Lived' Experience Anecdotal Evidence Other Theories Hunches

Lead to

363

Area of Enquiry

First ‘Slices-of-Data Categories and Their Properties

Theoretical Sampling

Adding Further Data Additional to Saturate Categories Additional Additional ‘Densification’ of the ‘Relations’ between ‘Slices-of-Data Categories

Grounded Theory

Relationships between Categories

Figure 1. Cycle of data collection and analysis in the grounded theory method [after Lehmann (2001) and Fernandez et al. (2002)].

Grounded theory views the process of theory generation as one of increasing the level of abstraction, range and scope of the theory. Generally speaking, there are three levels of theory in the grounded theory method.

Narrow concepts Seed concepts, which get the theory building process started, are of limited use and have the least range and scope. They are conceptual constructs themselves, although they are often no more than hunches, and have little, if any, empirical grounding. For example, Sarker et al. (2001) say that they started their research project by identifying aspects and concepts from their own backgrounds that could be brought to bear in their theorizing about virtual collaboration while guarding against becoming captive to any particular literature. They describe this first stage as one where they informally interacted with the data.

Substantive theories Theories that have been generated from within a specific area of enquiry using grounded theory methods are termed ‘substantive’ theories. They apply to the substantive area of enquiry, but are independent of and beyond the data analysed and the incidents observed


364

C Urquhart et al.

3. Formal Theories

2. Substantive Theories Categories Categories

Relations

Categories and Their

1. Narrow Properties Concepts

Figure 2. The progression of theory development in the grounded theory methodology (adapted from Lehmann, 2001).

(Glaser & Strauss, 1967). For example, Orlikowski (1993) developed a substantive theory relating to the use of CASE (Computer Aided Software Engineering) tools in organizations. She says that the concepts developed were intimately related to (because they were derived from) the arena of actual CASE tools adoption and use. At the same time, however, the theoretical framework was ‘sufficiently general to be applicable to a range of situations’ (Orlikowski, 1993, p. 335). Formal theories The highest level of abstraction in grounded theory is called a ‘formal theory’. Formal theories focus on conceptual entities (Strauss, 1987), such as organizational knowledge, organizational learning or collaborative work. Figure 2 depicts this hierarchy of theories. The general idea of using grounded theory is that as the researcher moves up the level of abstraction, the range and scope of the theory increases. Glaser & Strauss (1967) suggest that in order to generate formal theory, a comparative analysis should be made among different substantive theories that fall within a particular substantive area and by comparing substantive theoretical ideas from many different cases. Substantive theory can be used as a springboard towards formal theory by providing initial direction in developing conceptual categories. Glaser (1978) makes the point that constant comparative analysis can generate both substantive and formal theory. He further contends that in any study each type of theory can shade into the other. BUILDING THEORY IN THE INFORMATION SYSTEMS DISCIPLINE

Having examined the nature of grounded theory, its history, foundations and how a grounded theory is generated, we now turn our attention to how grounded theory can be applied specifically in the discipline of information systems.



365

The role of theory in information systems has commanded attention in recent years. For example, Gregor (2006) proposes a taxonomy of theory in information systems that includes the following categories – theory for analysing, theory for explaining, theory for predicting, theory for explaining and predicting, and theory for design and action. We suggest that grounded theory has the capability to generate theory that exists in all these categories because it contains the essential building blocks of any theory – constructs in the form of categories and relationships between those constructs in the form of theoretical coding. Dey (1993) uses the useful analogy of a wall of theory building – the categories are the bricks, the relationships the mortar between the bricks – but importantly, the emergent theory is informed by the research objectives. Gregor (2006) bemoans the use of vague terms for causality, like associated with or linked to, but from our perspective, it all depends for what purpose the theory is being developed – interpretive researchers find the search for causality limiting. What is certain is that because grounded theory has an emphasis on constructs and relationships, it is relatively easy to generate propositions relating to information systems phenomena that may – or may not – be testable. One well-known characteristic of the information systems research domain is the use of many theories borrowed from other disciplines (Baskerville & Myers, 2002). Although importing theories from outside the discipline is often valuable, we suggest that grounded theory could be used to build theories from within the field itself. For example, Orlikowski & Iacono (2001) pointed out the lack of presence of the IT artefact in theorizing in the information systems discipline. We suggest that grounded theory might well be useful in developing theories about such phenomena. Our next section introduces a framework for theorizing in grounded theory studies, and discusses levels of theory (theory scope), and the degree to which concepts are developed (theory conceptualization). This framework is then used as a basis for guidelines for information systems researchers using grounded theory.

A FRAMEWORK FOR THEORIZING IN GROUNDED THEORY STUDIES

We have observed in our own grounded theory work and in that of others that two aspects are important for theorizing. These two aspects are the degree of conceptualization and theory scope. These two dimensions underline the grounded theory process of theory building – conceptualization that moves beyond mere description, and also considers relationships between categories, and pitching the theory scope at the appropriate level. The first axis – the degree of conceptualization, can be seen as relating to the process of building a grounded theory, and relates to the degree of analysis carried out. The second axis, theory scope, can be seen to relate to the outcome of building a grounded theory. A summary of the framework is shown in Figure 3. As the main purpose of using grounded theory is theory building, researchers should aim to develop theories of greater scope. The more the data analysis moves from description to theory, and the more the scope of the theory increases with the development of formal


366

C Urquhart et al.

More Formal Concepts

Theory Scope

Substantive Focus

Bounded Context

Less Description Interpretation

Less

Theory

More

Degree of Conceptualisation Figure 3. A framework for analysing grounded theory studies.

concepts, the better. Generally speaking, grounded theory studies should aim for the top right-hand corner of the figure. Of course, the framework shown in Figure 3 might also be useful for other research approaches, but we suggest it is especially applicable to grounded theory. The axes of the figure will now be explained in more detail.

Degree of conceptualization The first axis of our framework is the degree of data analysis, corresponding to the horizontal axis of Figure 3. A key objective of grounded theory research is to aim for greater and greater depth of analysis of the data (Glaser & Strauss, 1967). In the Glaser variant, the process of discovering grounded theory has three principal stages that successively increase the depth of analysis.

Description The first stage yields descriptions. Descriptions are the most basic of conceptual constructs, where analysis has not proceeded beyond identifying concepts at the level of ‘categories’. Categories may have detailed ‘properties’, which are usually arrived at through a process of open coding.

Interpretation The second stage is the interpretation of categories and properties. Selective coding is employed to refine conceptual constructs that can help explain whatever interaction occurs between the descriptive categories (Glaser, 1978). This clear aetiological focus aims to



367

understand and explain one or more specific areas under investigation. At this point, the research problem becomes more refined, as aspects of the research problem become apparent through selective coding. Theoretical coding The third stage, theoretical coding, results in the formulation of a theory. The aim is to create inferential and/or predictive statements (sometimes in the form of hypotheses) about the phenomena. This is achieved by stipulating explicit relationships between individual interpretive constructs – these relationships can be associations or influences, or can be causal. The system of inferences covers the whole of the area under investigation. Without theoretical coding, there is no theory, as relationships between constructs have not been considered. Analytic memos, where relationships between categories are considered, are invaluable at this stage and assist the theoretical coding. There are a number of options open to the grounded theorist when considering relationships between categories. Glaser (1978) suggests 18 theoretical coding ‘families’. Strauss & Corbin (1990) suggest a coding paradigm. Clearly, how much attention is paid to the precise nature of the association between constructs is critical to theorizing. Theory scope The second dimension of the framework is that of theory scope. According to the tenets of grounded theory, the primary objective of the method is to develop theories of greater and greater scope (Glaser & Strauss, 1967, e.g. Dey, 1999). Bounded context Seed concepts within a ‘bounded context’ represent theory with the narrowest scope. Seed concepts, bounded by their immediate context within a specific area of inquiry, are often little more than hunches. These seed concepts have little empirical base; they simply represent formal postulates of the researcher’s hunches from ‘lived experience’, from anecdotal evidence that the researcher has about the field of enquiry, or even from limited, exploratory fieldwork. Substantive focus Theories with a ‘substantive focus’ are wider than theories within a bounded context. A substantive theory extends its predictive and explanatory power to the specific set of phenomena from where it was developed. This kind of theory is no longer simply based on seed concepts but has been developed by the rigorous application of grounded theory procedures. A substantive theory has significant empirical support. Formal concepts The widest form of grounded theory that can be developed is a formal theory that uses formal concepts. A formal theoretical construct applies to the conceptual area that it has been


368

C Urquhart et al.

developed for, which usually spans a set or family of several substantive areas. For example, a formal theory in information systems would apply to many different kinds of situations, systems and organizations (e.g. a theory regarding the implementation of information systems in general). While we have not found any instances of formal theories built from the use of the grounded theory method in the information systems discipline, this has been achieved in the social sciences – for instance, Biernacki’s (1986) theory of identity transformation. Clarke (2005) gives some useful strategies for formulating theories at the meso and macro level of analysis. For a grounded theorist, there are thus three levels of theory. The promise of grounded theory is that it can help a researcher to produce theories of greater and greater scope. Indeed, the main objective of a grounded theorist is to move up along the left axis as much as possible. Although our framework has two dimensions, we acknowledge that both axes are closely related to each other. We developed the framework primarily as a device to clarify what good grounded theory might look like. As we stated earlier, the main objective of grounded theory is to develop theory that is grounded in data systematically gathered and analysed. Obviously, the extent to which the data is analysed and conceptualized impacts the scope of the theory developed. Ideally, a researcher using grounded theory should attempt to move from the bottom left quadrant to the top right quadrant as much as possible (as per the diagonal arrow in Figure 3). The next section suggests guidelines for grounded theory studies in information systems using the two axes of Figure 3.

GUIDELINES FOR GROUNDED THEORY STUDIES IN INFORMATION SYSTEMS

A key question that this paper seeks to address is: ‘How can the grounded theory method be leveraged to build theory in information systems?’ This section attempts to answer this question by proposing guidelines for the conduct and evaluation of grounded theory studies in information systems. These guidelines are oriented towards building theory in our field, and are summarized in Table 1. The guidelines build on the two axes of the framework identified in the fourth section, conceptualization and theory scope. The first three guidelines address how the researcher might achieve the degree of conceptualization necessary to build a good theory through analytic mechanisms, such as constant comparison. These guidelines can also be seen as relating to the process of theory building. The final two guidelines give assistance with the issue of theory scope by giving guidance on the level of theory and how it might be integrated with the extant literature, an important aspect of theory building. Thus, these last two guidelines deal with the theory that is the outcome of the first three stages.

Constant comparison Constant comparison has been described as core to the grounded theory method (Charmaz, 2006). Constant comparison is the process of constantly comparing instances of data that you have labelled as a particular category with other instances of data in the same category to see



369

Table 1. Guidelines for grounded theory studies in information systems 1 Constant comparison

Constant comparison is the process of constantly comparing instances of data labelled as a particular category with other instances of data in the same category. Constant comparison contributes to the development of theory by exposing the analytic properties of the codes and categories to rigorous scrutiny. This guideline for data analysis encourages researchers to be both rigorous and theoretical (Charmaz, 2006).

2 Iterative conceptualization

This guideline suggests that researchers should increase the level of abstraction and relate categories to each other through a process of iterative conceptualization. In grounded theory, this is done using theoretical coding. The relationships between categories can be of many different types, not just causal. Theoretical coding contributes to an understanding of relationships between the concepts or factors of a theory. Theoretical memos are also very important to the development of theoretical coding and the whole process of iterative conceptualization.

3 Theoretical sampling

This guideline stresses the importance of deciding on analytic grounds where to sample from next in the study. Theoretical sampling helps to ensure the comprehensive nature of the theory, and ensures that the developing theory is truly grounded in the data.

4 Scaling up

This guideline suggests how a researcher might counter what is said to be a common problem in grounded theory viz. the production of a low level theory, which is then hard to relate to the broader literature. Scaling up is the process of grouping higher-level categories into broader themes. Scaling up contributes to the generalizability of the theory.

5 Theoretical integration

This guideline helps the researcher deal with what we think is an obligation of the grounded theorist – theoretical integration. Theoretical integration means relating the theory to other theories in the same or similar field. It is the process of comparing the substantive theory generated with other, previously developed, theories. This principle contributes to theoretical integration in the discipline and could help in the generation of formal theories.

if these categories fit and are workable (Urquhart, 2001). Charmaz (2006) makes two points about constant comparison. First, making comparisons between data, codes and categories advances conceptual understanding because of the need to expose analytic properties to rigorous scrutiny. Second, it makes the analysis more explicitly theoretical by asking ‘What theoretical category are these data an instance of?’ A legitimate question to be asked here is, should researchers code at the word and sentence level? Is it necessary to code at such a low level? From our perspective, the answer is a qualified yes, depending on the phenomena investigated. Such low-level coding is appropriate for interactional studies, because it means that the data are examined minutely, and just as importantly, ‘lived with’ for a long time. However, where the unit of analysis is the organization, word and sentence level coding may not always be as fruitful. That said, the insights that low-level coding affords cannot be underestimated, for it is in this way that the grounded theory method provides a chain of evidence like no other approach. The constant comparative method means that there are always dozens of instances to support the theory that is produced.


370

C Urquhart et al.

Constant comparison encourages the researcher to be both rigorous and theoretical (Charmaz, 2006). Both Orlikowski (1993) and Hughes & Howcroft (2000) are good examples of the use of constant comparison.

Iterative conceptualization One unique aspect of grounded theory is what we have chosen to call iterative conceptualization – where theory is built in an iterative fashion by using theoretical coding, focusing particularly on relationships between categories. These relationships can be of many kinds, causal relationships being one of many options. One of the interesting paradoxes about grounded theory is that, at first glance, it offers well-signposted procedures for theory building for the novice (Urquhart, 1997). Yet we notice that many researchers get into difficulties at this point, as theory building is an essentially creative process and cannot be achieved by following procedures alone. Hence, the strengths of the method can only be truly leveraged if both qualities of the method – systematic procedures and an iterative approach to conceptualization – are fully employed. Iterative conceptualization is the plank on which theory generation is based. A mechanistic application of coding stages will not yield the desired results in terms of theory. The researcher using grounded theory needs to be alert to intuition and to think beyond labels for the data. In terms of doing iterative conceptualization, researchers have suggested a number of alternatives. There are the coding stages of Strauss & Corbin (1990) (open coding, axial coding, selective coding), the coding stages of Glaser (1992) (open coding, selective coding, theoretical coding) or the coding stages of Charmaz (2006) (open coding, focused coding, axial coding, theoretical coding). Whichever coding stages are used, the key thing is that all stages are followed to allow adequate conceptualizations, which are the basis of a formed theory. Miles & Huberman (1994) give a useful set of characterizations about codes that are of assistance when assessing the data analysis component of grounded theory studies in information systems. They describe three types of codes that can be equated to analytic level: descriptive codes – attributing a class of phenomena to a segment of text, interpretive codes – where meaning is attributed with reference to context and other data segments, and pattern (or linked) codes – inferential and explanatory codes that describe a pattern. Clearly, it is desirable that the researcher reaches the third stage, that of inferential and explanatory codes. Axial coding (Strauss & Corbin, 1990) or theoretical coding (Glaser, 1978) are essentially about relationships between categories – the very essence of theory building. Theoretical coding contributes to an understanding of relationships between the concepts or factors of a theory. In our experience, it is in defining the relationships between categories that novice researchers often struggle to really achieve depth of theory. Establishing such relationships can be assisted by the coding paradigm of Strauss & Corbin (1990) and/or with Glaser’s (1978) coding families, which give many options for theory building, including causal relationships, and generation of hypotheses. Other options include considering limit, range, intensity, intent, aspects, types, dimensions, mutual interactions, ends and goals, clusters, and agreements.



371

During this stage, it often becomes clear that some categories are properties of others, and as thinking sharpens, category names often reflect analytic thinking as opposed to simply describing the phenomenon. Iterative conceptualization thus helps to answer important theoretical questions concerning ‘what’ and ‘why’. Whetten (1989) says that the ‘what’ in a theory justifies the selection of factors and the proposed (causal) relationships. The ‘why’ in a theory attempts to explain why the factors are behaving the way they do. This aspect of a theory supplies the plausible, cogent explanation for ‘why we should expect certain relationships in the what and how data’ (Whetten, 1989, p. 491). Theoretical memos (Strauss, 1987; Charmaz, 2006) are a further aid to iterative conceptualization, as the writing of a memo creates a formal space in which the researcher can reflect on the emerging theory. Examples of theoretical memos are rare in information systems, but not in the social sciences (Charmaz, 2006). Examples of theoretical memos can be found in Urquhart (1997; 2001).

Theoretical sampling Theoretical sampling is deciding on analytic grounds where to sample from next (Glaser & Strauss, 1967), and is an important aspect of grounded theory. Theoretical sampling can occur at the group level (considering similar and different data sets) and at the category level through ‘slices of data’ (Glaser & Strauss, 1967). Through successive sampling according to the emergent theory (Glaser, 1992), the research questions gradually become more refined, as dimensions of the research problem become clearer through analysis (Dey, 1993). If the researcher is guided by the emergent theory when collecting data, then there is very little chance of the researcher imposing preconceived notions on the data. Theoretical sampling also means that there is a focus on the development of research questions. Orlikowski (1993) employed theoretical sampling in her study of organizational change and CASE tools. Charmaz (2006) suggests that theoretical sampling assists delineation of category properties, relationships between categories, to saturate categories, to clarify relationships between categories, to distinguish between categories, and to follow hunches about categories. Theoretical sampling is one of the foundations of grounded theory method – it enables both a focus on the developing theory and ensures that the developing theory is truly grounded in the data. Theoretical sampling can also be used to extend the scope of the generated theory. Theoretical sampling is a key element of the method and the single most important contributor to the ‘fit’ of a theory. Without theoretical sampling and the constant comparison and assessment of the contribution achieved by new slices of data, it will be impossible to establish how ‘saturated’ the theory is. In fact, theoretical sampling is the single most important assurance that a theory ‘works’, i.e. explains ‘what is actually going on’ (Glaser & Strauss, 1967, p. 35). Most of the theoretical sampling in the information systems research literature is rather weak and seems to have taken the path of ‘more-of-the-same’, which mainly serves to confirm the properties of existing categories and can freeze the current conceptual level. ‘Same-datagroup’ is, however, but one of several options, some of them explicitly designed to extend


372

C Urquhart et al.

scope and depth of the theoretical constructs created. Glaser & Strauss (1967, pp. 49–60) and Glaser (1978, pp. 36–54) provide a detailed discussion of such theoretical sampling strategies. This sampling for data that could enhance the theory is not only needed during a specific research project, but also should be carried on even after a first cohesive theoretical construct has been established. This is for two reasons: first, it enables maximum ‘fit’ of the theory by keeping it up-to-date with changing circumstances; and second, it facilitates the extension of the theory’s substantive limits. Theoretical sampling ensures the comprehensive nature of the theory. Deciding on which categories are ‘core’ categories and selectively coding until saturation is reached also provides a comprehensive theory that is well grounded in the data. Because each category needs to be ‘saturated’, that is, well represented by many instances in the data, the theory generated is also parsimonious. Whetten (1989) suggests that comprehensiveness and parsimony are two important characteristics of theory.

Scaling up Our collective experience with the grounded theory method tells us that first-time users tend to get overwhelmed at the coding level. The attention to word- and sentence-level coding, while giving rich insights to the researcher, naturally focuses the mind on the detail. However, we believe it is important for researchers at some stage to try to rise above the detail in order to consider the bigger picture. One simple mechanism that we have used successfully is that of grouping high-level categories into larger, broader themes. It is then much easier to relate these to competing theories. The desired level of abstraction can be achieved by coding around one or two core categories or themes (Glaser, 1978; 1992; Strauss, 1987). In practice, we find that people end up with far more than one or two core categories, particularly if they start with word- or sentence-level coding. This may be because the phenomenon being studied is not necessarily a process, and may have many different and distinct elements. A more compelling general reason is that the bottom up derivation of the generated theory makes it difficult to think abstractly. The very strength of grounded theory – its unique tie to the data – may also be in fact the Achilles heel of the method. Thus, grouping high-level categories (which may or not be named core categories) into higher-level core categories (or themes) is a very useful practice to scale up the substantive theory. A similar tactic is to generate propositions. Glaser makes some useful suggestions for scaling up a theory (Glaser, 1978). First, the rewrite method, where the theory is rewritten to omit specifics – so, for example, instead of talking about the strategies used by analysts when talking to their clients, one could talk about the strategies used by professionals when dealing with their clients. While there are no more data points sampled, it nevertheless provides a starting point for increasing the level of conceptualization. Second, the level of conceptualization can be raised by comparing it to the data from other substantive theories.



373

Hence, scaling up contributes to answering the ‘who, where and what’ aspects of a theory. Whetten suggests these aspects are very important in determining the temporal and contextual factors that set the limit on the theory’s range i.e. determine how generalizable the theory is (Whetten, 1989).

Theoretical integration Like any other theory, a grounded theory needs to be put into the context of other theories in the field. One of the potential advantages of the grounded theory method for information systems researchers is the obligation (Strauss, 1987, p. 282) to engage with theories outside the discipline. Weber (2003) suggests more scrutiny of high quality exemplars from other disciplines. We believe that the use of grounded theory, because of its rigorous approach to theory building, makes this kind of scrutiny possible. Glaser (1978) suggests that the substantive theory can be analysed by comparing it with other substantive theories in the area. Glaser also suggests that formal models of process, structure and analysis may be useful guides to integration. Hence, in the field of information systems, meta-theories such as structuration theory (Orlikowski & Robey, 1991; Walsham, 2002) or actor–network theory (Walsham, 1997) may be a useful lens through which to view the emergent theory. Glaser (1978) also makes the point that context is necessarily stripped away as one moves toward a formal theory, and that comparative analysis is used to compare conceptual units of a theory, as well as data. In the next section, we illustrate how the guidelines of constant comparison, iterative conceptualization, theoretical sampling, scaling up and theoretical integration can be applied to grounded theory studies in information systems.

APPLYING THE GUIDELINES

In this section, we illustrate the usefulness of the guidelines by applying them to three grounded theory studies in information systems. The three studies are as follows: 1 Orlikowski’s (1993) study of the use of CASE tools in two organizations; 2 Urquhart’s (2001) study of the dialogue between a systems analyst and client in one of six case studies; and 3 Lehmann & Gallupe’s (2005) analysis of the use of information systems in three multinational companies across multiple locations.

Orlikowski (1993) Orlikowski (1993) investigates the adoption and use of CASE tools in organizations. Most of the data comes from interviews with people from two organizations. The data analysis was carried out following the Strauss and Corbin stages of coding – open coding, axial coding


374

C Urquhart et al.

and selective coding. Orlikowski uses the terms concepts, properties and relations, as used in Glaser & Strauss (1967). Concepts are grouped in categories, and those categories related together in a framework. There are six major categories in the framework, substantially more than the one or two core categories suggested by Glaser and Strauss as the basis of a theory. Constant comparison was used and is explained as constant comparison across types of evidence to control the conceptual level and scope of the emerging theory (Orlikowski, 1993). Orlikowski also used iterative conceptualization to draw out ‘multiple sources of loops of causation and connectivity’ and to identify patterns in the process of change. Connections were made between subcategories using the axial coding technique (Strauss & Corbin, 1990). It is not clear whether the Strauss & Corbin (1990) coding paradigm of phenomena, causal conditions, context, intervening conditions, action and interactional strategies, and consequences was used to assist the coding. Conditions do appear in the theory, leading us to wonder if Orlikowski chose to be informed by, rather than apply, the coding paradigm. The relationships between three categories (Environmental Context, Organizational Context and Information Systems Context) and two other categories (Conditions for Adopting and Using CASE Tools and Adopting and Using CASE tools) are very carefully delineated and discussed in the paper. It appears that theoretical sampling (deciding on analytical grounds where to sample from next) was used in the later stages of data collection. Data collection was overlapped with analysis, and the author cites Eisenhardt (1989) as pointing out that this is useful for theory building. The author also says that in the site selection, the two cases were chosen for their similarities, as well as their differences. Glaser & Strauss (1967) say that minimizing differences among comparison groups increases the possibility of collecting similar data, but will also help in spotting important differences. Scaling up is also evident in the paper, as eight categories are generated in total. Three are grouped into the Institutional Context, and three into Strategic Conduct in Adopting and Using CASE tools. There are also three categories on Systems Developer Reactions to CASE tools that are not linked to the preceding framework, but discussed separately under the heading of ‘Implications for Systems Development Practice’. Theoretical integration is achieved by discussing the resultant theory in the context of other theories on radical change and distinctions in types of innovation, in particular the classification of incremental and radical types of innovation used in the innovation literature.

Urquhart (2001) Urquhart (2001) explores three case studies of analyst–client interaction and discusses three themes from three case studies in the context of Boden’s (1994) theory of organizational agendas. Constant comparison is not mentioned in the article, though both the individual characteristics of the analysts and clients and the results from the three cases are explicitly compared.



375

The sequence of coding is explained in detail, including how different data sources were treated in the coding process. Similarly, theoretical sampling is not evident in the paper (though it did take place at the category level as the author can attest). By contrast, iterative conceptualization and scaling up are clearly evident in the explanation of grouping grounded theory concepts, via an intermediate unit of analysis, a conversational topic. In terms of iterative conceptualization, there are a number of different relationships postulated between the themes Organizational Context, Issues to Be Discussed and Professional Relationships. Scaling up of the theory is achieved by grouping categories into themes using the aforesaid intermediate unit of analysis – the conversational topic. Actual categories are not mentioned. Theoretical integration is achieved by discussing the emergent theory, in particular the relationships between the three themes, in the light of Boden’s (1994) theory of organizational agendas arising from interaction. The findings illustrate how an organizational agenda may start from an interaction. Also, the reflexive nature of this relationship is discussed.

Lehmann & Gallupe (2005) Lehmann & Gallupe (2005) examine three multinational companies and put forward a theoretical framework concerned with the structure of international information systems. The constant comparative method is explicitly mentioned, and data collection overlapped with data analysis and theory building. Iterative conceptualization was carried out using Weick’s (1979) cause–effect loops between categories. These cause–effect loops were built on theoretical coding between categories and are very extensive. There are also relationships posited between Global Standards, Information Systems Initiatives, Strategic Migration, Autonomy and Information Systems by force – all categories from the grounded theory analysis. Theoretical sampling was used at the level of text, the case and across cases. Scaling up of theory is evident – a force-field diagram showing tensions between territorial forces and functional forces appears to be grounded in the categories. Theoretical integration is achieved by relating the substantive theory to other theories. The international information systems architecture model is systematically related to global strategy literature. Examples of cyclical movements are mentioned from many diverse business research fields, indicating that the authors are both raising the level of conceptualization by comparing the theory to data from other substantive theories, and using theoretical sampling to a great extent. While the grounded nature of the theory is evident, and relationships between concepts clearly explained, this paper does not explicitly show the route from categories to larger concepts. Neither does the paper explicitly discuss the coding, except in general terms. Again, this illustrates a significant problem for grounded theory studies – in the space afforded by a journal article, the author may have to make a choice between explaining the theory and explaining the chain of evidence that led to that theory. Our overall assessment of all three articles using the five guidelines for grounded theory studies is shown in Table 2.


376

C Urquhart et al.

Table 2. Overall assessment of the three grounded theory studies Lehmann & Gallupe (2005)

Orlikowski (1993)

Urquhart (2001)

1 Constant comparison

Constant comparison is explicitly mentioned. The first case was systematically contrasted with the second case

Constant comparison is not explicitly mentioned in the paper, but is evident in cross-case comparison.

Constant comparison is explicitly mentioned in the paper.

2 Iterative conceptualization

Connections were made between subcategories using axial coding. Interactions between 3 key concepts and the context were carefully outlined.

Situational links between themes are made, and these links are based on lower level category relationships

The categories have correlational and causal linkages, and hypotheses are formed

3 Theoretical sampling

Later stages of data collection were directed by emerging concepts, and data collection was overlapped with analysis. Sites chosen for similarities rather than differences.

Theoretical sampling is implied, but not explicitly mentioned. No overlapping data collection and analysis.

Theoretical sampling carried out at the level of text, case and across cases. Overlapping data collection and analysis.

4 Scaling up

Scaling up is evident, as each key concept contains a number of categories.

Scaling up is evident as categories within an intermediate unit of analysis, conversational topic, are grouped into themes.

Scaling up is not explicit, but appears to have been done to some extent.

5 Theoretical integration

The resultant theory is discussed in the context of other theories on radical change.

The resultant theory is related to Boden’s (1994) work on organizational agendas.

The resultant theory is related to Lewin’s (1952) force field concept.

As can be seen in the table, in two out of the three papers analysed, constant comparison was explicitly mentioned. In all three papers, iterative conceptualization was evident. Theoretical sampling was evident in two of the papers – the other paper did a series of ‘one-shot’ case studies so no overlapping data collection and analysis occurred. This is perhaps typical of grounded theory studies that start out using grounded theory as a method of analysis only. Scaling up is very obvious in two of the papers, and it may be that in the third paper, it has occurred but not mentioned explicitly. All three papers relate their emergent theory to larger theories. Our concluding section discusses the ramifications of our guidelines and their potential contribution to the use of grounded theory in information systems.



377

DISCUSSION AND CONCLUSIONS

Over the past decade, many information systems researchers have started to use the grounded theory method. While we welcome this growing interest in the method, we believe it is an opportune time to question whether information systems researchers have been using this method to its full potential. It would seem that many information systems researchers, like those in other disciplines, have used the grounded theory method mostly as a way of coding qualitative data (Becker, 1983; Benoliel, 1996; Bryant et al., 2004; Urquhart, 2007). This use of grounded theory, while appropriate in some cases, suggests to us that the primary purpose for which grounded theory was developed – to generate theory – is being neglected. Grounded theory is not just a coding technique, but offers a comprehensive method of theory generation. There have been calls for information systems researchers to focus more on theory development (Watson, 2001; Weber, 2003). We have suggested that one possible way to answer this call is to use grounded theory to help generate theories related to information systems phenomena. Hence, the key question that this paper has sought to address is: ‘How can the grounded theory method be leveraged to build theory in information systems?’ We have answered this question by suggesting guidelines for grounded theory studies in information systems. The guidelines are oriented towards increasing the degree of conceptualization and theory scope in grounded theory studies. Our intention is to raise the bar for grounded theory studies in information systems, such that all information systems researchers who use grounded theory might aim to increase the degree of conceptualization and theory scope in their research as much as possible. Our suggested guidelines draw attention to a few key features of the grounded theory method. First, constant comparison is at the heart of the method. Constant comparison helps to ensure that the categories and the resulting theory are properly grounded. The idea of one or two core categories or themes helps focus the theory. Iterative conceptualization is also fundamental to the method. The dynamic interplay between analysis and data collection – where relationships are built between concepts in an iterative manner – is one of the features that distinguishes grounded theory from most other qualitative research methods. Theoretical sampling increases the relevance and density of the theory, while scaling up helps to increase the level of abstraction. Theoretical integration, where the generated theory is related to other theories, has the potential to help bring disparate theory building efforts together. While Lee & Baskerville (2003) counsel against the ‘uniformity of nature assumption’ in information systems theory building efforts, we would suggest that the utilization of grounded theory for theory building in information systems would in fact increase our discipline’s engagement with diverse theories from other fields. We have suggested that the guidelines, while potentially helping to improve the conduct of grounded theory studies in information systems, can also be used for post hoc evaluation. We evaluated three such information systems articles, and found that all three exhibited the five guidelines for grounded theory studies to some extent, but most guidelines were implicit rather than explicit. There was some inconsistency in application, and some guidelines were empha-


378

C Urquhart et al.

sized more than others. For example, constant comparison tended to be a less visible part, especially in one of the studies. In all of the articles, scaling up was evident, but none of the authors explained how this was achieved. In fact, in one of the studies, scaling up was not mentioned at all, although the findings would suggest that it must have been performed to some extent. All three articles were exemplary in their attempts at theoretical integration with previously existing theories. In our opinion, the article by Orlikowski (1993) remains the high-water mark for theorizing in information systems using grounded theory. All five guidelines are clearly evident in this paper. The article pays great attention to the relationships between concepts, exhibits iterative conceptualization and systematically explores those relationships. It also provides more of a chain of evidence than the other two papers, although the author does not explain how scaling up was achieved. It is interesting to speculate why this is the best paper of the three. It may be that a less restrictive word limit for journal articles may be partly responsible for the depth and excellence of Orlikowski’s theorizing: MIS Quarterly allows for the publication of longer papers than most other journals. We believe that our evaluation of these articles using the five guidelines demonstrates how grounded theory can be leveraged to build theory in information systems. The guidelines emphasize the key distinguishing features of grounded theory and suggest how theory building efforts that use grounded theory in information systems might be improved. One question that can legitimately be asked is whether more use of grounded theory in information systems would simply result in more unrelated theories. We acknowledge that this is a possibility, but one way to counter this would be to consider extended use of theoretical sampling. Extending theoretical sampling points to a way of increasing the explanatory power of grounded theories: it allows information systems researchers to build on each other’s work with studies that complement or extend earlier work. These studies can be carefully designed to extend either the ‘fit’ of the theory, thereby increasing its scope, or to improve the way the theory ‘works’ by modelling more, and more specific, linkages and relationships between the ‘objects’ (i.e. concepts and constructs of the existing theory). Other possibilities would be the extension of the theory’s original substantive area. This brings us to the issue of collaboration. Students of Anselm Strauss were encouraged to code collaboratively, and it has also been our experience that collaborative coding results in a stronger theory. Follow-on studies using generated theories could be carried out in collaborative arrangements, perhaps by forming virtual teams. One of the grounded theorist’s most powerful tools, the theoretical memo, lends itself naturally to email communication. One example of Internet collaboration is the Forum for researchers on Glaser’s Grounded Theory Institute web page (Glaser, 2008) Collaboration could also be done through the exchange of data sets from text analysis software applications such as NUDIST, nVivo or ATLAS ti. In closing, we caution that the five guidelines we have suggested should not be used mechanistically. Although the grounded theory method is sometimes seen as rather formulaic and overburdened with rules (Melia, 1996; Kendall, 1999), we would like to stress that the application of the guidelines requires considerable creative thought (cf. Klein & Myers, 1999).



379

The use of the guidelines does not obviate the need for intellectual effort on the part of the researcher. In summary, we have suggested guidelines for grounded theory studies in information systems. These guidelines are oriented towards increasing the degree of conceptualization and theory scope in grounded theory research projects. Our hope is that these guidelines will to help to raise the quality and aspirations of grounded theory studies in information systems. If grounded theory is used to its full potential, we believe that we may well see much more theory development in the information systems arena.

REFERENCES Annells, M.P. (1996) Grounded theory method: philosophical perspectives, paradigm of inquiry, and postmodenism. Qualitative Health Research, 6, 379–393. Baskerville, R.L. & Myers, M.D. (2002) IS as a reference discipline. MIS Quarterly, 26, 1–14. Becker, P.H. (1983) Common pitfalls in published grounded theory research. Qualitative Health Research, 3, 254–260. Benoliel, J.Q. (1996) Grounded theory and nursing knowledge. Qualitative Health Research, 6, 406–428. Biernacki, P. (1986) Pathways from Heroin Addiction: Recovery without Treatment. Temple University Press, Philadelphia, PA, USA. Boden, D. (1994) The Business of Talk. Polity Press, Cambridge, UK. Bryant, A. (2002) Re-grounding grounded theory. Journal of Information Technology Theory and Application, 4, 25–42. Bryant, A., Hughes, J., Myers, M.D., Trauth, E.M. & Urquhart, C. (2004) Twenty years of applying grounded theory in information systems: a coding method, useful theory generation method, or an orthodox positivist method of data analysis? In: Information Systems Research: Relevant Theory and Informed Practice, Kaplan, B., Truex, D.P., Wastell, D., Wood-Harper, A.T. & DeGross, J.I. (eds), pp. 649–650. Kluwer Academic Publishers, Norwell, MA, USA. Cecez-Kecmanovic, D., Klein, H.K. & Brooke, C. (2008) Exploring the critical agenda in information systems research. Information Systems Journal, 18, 123–135. Charmaz, K. (2005) Grounded theory in the 21st century: a qualitative method for advancing social justice research. In: The Handbook of Qualitative Research, Denzin, N.K. & Lincoln, Y.S. (eds), pp. 507–535. Sage, Thousand Oaks, CA, USA.

Charmaz, K. (2006) Constructing Grounded Theory: A Practical Guide through Qualitative Analysis. Sage Publications, Thousand Oaks, CA, USA. Clarke, A. (2005) Situational Analysis: Grounded Theory after the Post Modern Turn. Sage Publications, Thousand Oaks, CA, USA. Dey, I. (1993) Qualitative Data Analysis. Routledge, London, UK. Dey, I. (1999) Grounding Grounded Theory: Guidelines for Qualitative Inquiry. Academic Press, San Diego, CA, USA. Eisenhardt, K.M. (1989) Building theories from case study research. Academy of Management Review, 14, 532– 550. Elliott, N. & Lazenbatt, A. (2005) How to recognise a ‘quality’ grounded theory research study. Australian Journal of Advanced Nursing, 22, 48–52. Fernandez, W., Lehmann, H.P. & Underwood, A. (2002) Rigour and relevance in studies of IS innovation: a grounded theory methodology approach. In: Information Systems and the Future of the Digital Economy, Proceedings of the 10th European Conference on Information Systems, Wrycza, S. (ed.), pp. 110–119. 6–8 June 2002, Gdañsk, Poland. Glaser, B.G. (1978) Theoretical Sensitivity: Advances in the Methodology of Grounded Theory. The Sociology Press, Mill Valley, CA, USA. Glaser, B.G. (1992) Emergence vs. Forcing: Basics of Grounded Theory Analysis. Sociology Press, Mill Valley, CA, USA. Glaser, B.G. (ed.) (1995) Grounded Theory: Volumes 1 and 2. Sociology Press, Mill Valley, CA, USA. Glaser, B.G. (1998) Doing Grounded Theory: Issues and Discussions. Sociology Press, Mill Valley, CA, USA.


380

C Urquhart et al.

Glaser, B.G. (1999) The future of grounded theory. Qualitative Health Research, 9, 836–845. Glaser, B.G. (2001) The Grounded Theory Perspective: Conceptualization Contrasted with Description. Sociology Press, Mill Valley, CA, USA. Glaser, B.G. (2008) The Grounded Theory Institute. The official site of Dr. Barney Glaser and classic grounded theory. [WWW document]. URL http://www. groundedtheory.com/ [accessed 11 November 2008]. Glaser, B.G. & Strauss, A.L. (1967) The Discovery of Grounded Theory: Strategies for Qualitative Research. Aldine Publishing Company, Chicago, IL, USA. Goulielmos, M. (2004) Systems development approach: transcending methodology. Information Systems Journal, 14, 363–386. Green, J. (1998) Grounded theory and the constant comparative method. British Medical Journal, 316, 1064– 1065. Gregor, S. (2006) The nature of theory in information systems. MIS Quarterly, 30, 611–642. Holton, J.A. (2007) The coding process and its challenges. In: The Sage Handbook of Grounded Theory, Bryant, A. & Charmaz, K. (eds), pp. 265–289. Sage, London, UK. Howcroft, D. & Hughes, J. (1999) Grounded theory: I mentioned it once but I think I got away with it. In: Information Systems – The Next Generation: Proceedings of the 4th UKAIS Conference, Brooks, L. & Kimble C. (eds), pp. 129–141. McGraw-Hill, Maidenhead, UK. Hughes, J. & Howcroft, D. (2000) Grounded theory: never knowingly understood. Information Systems Review, 1, 181–197. Kendall, J. (1999) Axial coding and the grounded theory controversy. Western Journal of Nursing Research, 21, 743–757. Klein, H.K. & Myers, M.D. (1999) A set of principles for conducting and evaluating interpretive field studies in information systems. MIS Quarterly, 23, 67– 93. Lee, A.S. & Baskerville, R.L. (2003) Generalizing generalizability in information systems research. Information Systems Research, 14, 221–243. Lehmann, H.P. (2001) Using grounded theory with technology cases: distilling critical theory from a multinational information systems development project. Journal of Global Information Technology Management, 4, 45–60. Lehmann, H.P. (2003) An object oriented architecture model for international information systems? An exploratory study. Journal of Global Information Management, 11, 1–18.

Lehmann, H.P. & Gallupe, B. (2005) Information systems for multinational enterprises – some factors at work in their design and implementation. Journal of International Management, 11, 28–49. Lewin, K. (1952) Field Theory in Social Science. Tavistock, London, UK. Lings, B. & Lundell, B. (2005) On the adaptation of Grounded Theory procedures: insights from the evolution of the 2G method. Information Technology & People, 18, 196–211. Lundell, B. & Lings, B. (2003) The 2G method for doubly grounding evaluation frameworks. Information Systems Journal, 13, 375–398. Madill, A., Jordan, A. & Shirley, C. (2000) Objectivity and reliability in qualitative analysis: realist, contextualist and radical constructionist epistemologies. British Journal of Psychology, 91, 1–20. Martin, P.Y. & Turner, B.A. (1986) Grounded theory and organizational research. The Journal of Applied Behavioral Science, 22, 141–157. Melia, K.M. (1996) Rediscovering glaser. Qualitative Health Research, 6, 368–373. Miles, M.B. & Huberman, A.M. (1984) Qualitative Data Analysis: A Sourcebook of New Methods. Sage Publications, Newbury Park, CA, USA. Miles, M.B. & Huberman, A.M. (1994) Qualitative Data Analysis: An Expanded Sourcebook. Sage Publications, Newbury Park, CA, USA. Myers, M.D. (1997) Qualitative research in information systems. MIS Quarterly, 21, 241–242. Orlikowski, W.J. (1993) CASE tools as organizational change: investigating incremental and radical changes in systems development. MIS Quarterly, 17, 309– 340. Orlikowski, W.J. & Iacono, C.S. (2001) Research commentary: desperately seeking the ‘IT’ in IT research – a call to theorizing the IT artifact. Information Systems Research 12, 121–134. Orlikowski, W.J. & Robey, D. (1991) Information technology and the structuring of organizations. Information Systems Research, 2, 143–169. Sarker, S., Lau, F. & Sahay, S. (2001) Using an adapted grounded theory approach for inductive theory building about virtual team development. The DATA BASE for Advances in Information Systems, 32, 38–56. Strauss, A. (ed.) (1987) Qualitative Analysis for Social Scientists. Cambridge University Press, Cambridge, UK. Strauss, A. & Corbin, J. (1990) Basics of Qualitative Research: Grounded Theory Procedures and Techniques. Sage Publications, Newbury Park, CA, USA.



Strauss, A. & Corbin, J. (1994) Grounded theory methodology – an overview. In: Handbook of Qualitative Research, Denzin, N.K. & Lincoln, Y.S. (eds), pp. 273– 285. Sage Publications, Thousand Oaks, CA, USA. Strauss, A. & Corbin, J. (eds) (1997) Grounded Theory in Practice. Sage Publications, London, UK. Urquhart, C. (1997) Exploring analyst-client communication: using grounded theory techniques to investigate interaction in informal requirements gathering. In: Information Systems and Qualitative Research, Lee, A.S., Liebenau, J. & DeGross, J.I. (eds), pp. 149–181. Chapman and Hall, London, UK. Urquhart, C. (2001) An encounter with grounded theory: tackling the practical and philosophical issues. In: Qualitative Research in IS: Issues and Trends, Trauth, E. (ed.), pp. 104–140. Idea Group Publishing, Hershey, PA, USA. Urquhart, C. (2007) The evolving nature of grounded theory method: the case of the information systems discipline. In: The Handbook of Grounded Theory, Charmaz, K. & Bryant, T. (eds), pp. 311–331. Sage Publishers, London, UK. Walsham, G. (1997) Actor-network theory and IS research: current status and future prospects. In: Information Systems and Qualitative Research, Lee, A.S., Liebenau, J. & DeGross, J.I. (eds), pp. 466–480. Chapman and Hall, London, UK. Walsham, G. (2002) Cross-cultural software production and use: a structurational analysis. MIS Quarterly, 26, 359–380. Watson, R. (2001) Research in information systems: what we haven’t learned. MIS Quarterly, 25, v–xv. Weber, R. (2003) Editor’s comments: theoretically speaking. MIS Quarterly, 27, iii–xii. Weick, K.E. (1979) The Social Psychology of Organizing. University of Michigan, Ann Arbor, MI, USA. Whetten, D.A. (1989) What constitutes a theoretical contribution? Academy of Management Review, 14, 490–495.

381

MIS Quarterly in 2007. She has a strong interest in qualitative data analysis, especially the use of grounded theory in information systems. Her current research interest is technology and social inclusion, particularly in developing countries. She is an Associate Editor for the Journal of Information Technology and Development and the International Journal of E-Politics, and is on the editorial board for the International Journal of Learning and Change. Her home page can be accessed at http:// staff.business.auckland.ac.nz/curquhart.

Cathy Urquhart is a Senior Lecturer in Information Systems at the Department of Information Systems and

Hans Lehmann is the Associate Professor for Electronic Business at Victoria University of Wellington, New Zealand. He is a graduate in psychology from the University of Vienna and the University of Natal and in business administration from the University of South Africa. His PhD in Information Systems was obtained from the University of Auckland. Hans looks back on 25 years of business experience with information technology, both in line management in banks and the manufacturing industry and as a consultant with Deloitte specializing in the management of information systems for multinational enterprises. In 1991, Hans changed careers and joined the University of Auckland, New Zealand, where he focused his research on the strategic management of global information technology. In 2003, he moved to Victoria University, where his current research interest is in the application of wireless technology in business. His research focuses strongly on qualitative enquiry, with the main emphasis on the use of grounded theory. Michael D. Myers is Professor of Information Systems and Head of the Department of Information Systems and Operations Management within the University of Auckland Business School, New Zealand. His research articles have been published in many journals and books. He won the Best Paper Award (with Heinz Klein) for the most outstanding paper published in MIS Quarterly (MISQ) in 1999. This paper has been cited over 1000 times and is the third most cited paper to appear in MISQ. He also won the Best Paper Award (with Lynda Harvey) for the best paper published in Information Technology & People in 1997. He currently serves as Senior Editor of Information Systems Research and as Editor of the ISWorld Section on Qualitative

Operations Management at the University of Auckland Business School, New Zealand. She has a PhD in Infor-

Research. He previously served as Senior Editor of MIS Quarterly from 2001–2005, as Associate Editor of Informa-

mation Systems from the University of Tasmania, Australia. She was named as one of Australia’s outstanding

tion Systems Research from 2000–2005 and as Associate Editor of Information Systems Journal from 1995–2000. He

teachers of computing in the Australian Campus Review in November 1996. She won the Outstanding Paper

also served as President of the Association for Information Systems in 2006–2007, and as Chair of the International

award in Information Technology and People in 1999. She won a Developmental Associate Editor Award for

Federation of Information Processing Working Group 8.2 from 2006–2008.

Biographies


PUTTING THE THEORY BACK INTO GROUNDED THEORY

Recommend Documents