using an adapted abstract - heim.ifi.uio.noheim.ifi.uio.no/~sundeeps/publicationsnew/sundeep... ·...

Using an Adapted Grounded Theory Approach for Inductive Theory Building About Virtual Team Development Suprateek Sarker Washington State University

Francis Lau University of Alberta,. Edmonton, Canada

Sundeep Sahay University of Oslo, Norway

Acknowledgements

The project leading to this paper was funded by the Learning Enhancement Envelope from the Ministry of Alberta Advanced Education and Career Development, the University Teaching Research Fund at the University of Alberta, and the Office of the Dean, School of Business and Public Management at the George Washington University.

An earlier version of this paper was presented at HICSS 2000. The authors would like to thank the HICSS 2000 reviewers and the mintrack chairs, especially Sajda Qureshi. They would also like to acknowledge the constructive comments and encouragement of Joe Valacich and Richard Orwig of Washington State University.

All three authors contributed equally to the paper.

Abstract

This paper outlines how the grounded theory methodology was adapted to develop a process model of collaboration in virtual teams. The data analysis was conducted using an adapted version of open coding, axial coding, and selective coding procedures offered by Strauss and Corbin (1990). In applying the grounded theory procedures, the objective was to stay true to the goals and spirit of each coding procedure, while modifying details of procedural steps that were too mechanistic or impractical. The paper develops a meta-theoretical framework through a synthesis of the data, the symbolic interactionist perspective, and structuration theory. This framework is an alternative to the "paradigm model" during selective coding of data.

ACM Categories: H4.3, K3.1, K.6.1, J.4, D.2.1

Keywords: qualitative methodology, grounded theory, coding, interpretive, virtual team development

Introduction This paper illustrates how the grounded theory methodology, suitably modified, can be used to develop theory about new forms of IT-enabled organizations such as "virtual teams" and associated phenomena in a systematic fashion. As such innovative forms of human organizations appear, there is a need to understand these organizational forms in order to evaluate and manage them effectively (Townsend et al., 1998; Warekentin et al., 1997).

There are two broad approaches to investigate aspects of new organizational forms: (1) testing in the new context (say, "virtuality") of existing theories on related aspects of traditional organizational forms synthesized with the researcher's intuitive understanding of the new forms or what is known from exploratory studies regarding these new forms, and (2) developing a theoretical understanding of the new forms that is grounded in the experiences of human subjects who are/have been members of such forms. The latter approach is more useful in situations where the new organizational forms are so novel or different from traditional forms that theories and insights on traditional forms cannot be easily translated or directly extended to understand or explain phenomena pertaining to the new forms. A deductive

38 The DATA BASE for Advances in Information Systems - Winter 2001 (Vol. 32, No. 1)

approach may be useful once a certain level of inductive understanding of data has been attained.

The goal of the larger project (from which this paper is derived) was to investigate virtual teams, and the investigation used the latter (i.e., predominantly inductive) approach. The reason for the approach is that until the nature of virtual teams and the phenomena surrounding them are understood in some depth, virtual teams need to be studied, not deductively, based on propositions existing in the teams/group/traditional organization behavior literature (e.g., Jarvenpaa et al., 1998), who tested traditional theories of "trust" in the context of virtual teams), but inductively, based on the collaborative experiences of virtual team members and the meanings they attribute to the virtual experiences. Of course, an inductive approach need not ignore existing literature and researchers' personal experiences - - the guiding principle in such an approach is to consciously avoid being driven by pre-conceptions while remaining true to the data (Trauth, 1997).

Of the different methodologies available for inductive theory building such as interpretive case study, ethnography, hermeneutics, ethnomethod- ology, and grounded theory, the grounded theory methodology was chosen for the following ma- sons:

• It emphasizes,, as much or more than all other inductive methodologies, the need for the researchers to be immersed in data, and the need to consciously guard against imposing a theory in a related substantive area that does not actually match the patterns in the data (Glaser & Strauss, 1967; Urquhart, 1997);

• It does not require the researchers to sus- pend or ignore all pre-existing theoretical knowledge (as in, for example, ethnometho- dology), but, instead, encourages the development/enriching of grounded theories by drawing upon (not driven by) broad theoretical approaches that are not in the same substantive area (Glaser, 1978);

• It draws on the strengths of both the positivist and interpretivist approaches I (Charmaz 2000). The grounded theory approach has

1 This is especially true regarding the version of grounded theory methodology described by Strauss and Corbin (1990), which was the methodological perspective used in this investigation.

been characterized as "interpretive" (Orlikowski, 1993), in that: (1)it uses qualitative and unstructured data that represent the subjective understanding of actual members of the new organizational forms (Strauss & Corbin, 1990); (2) it involves subjective sampling and analysis techniques (Flick, 1998); and (3) theory-building strategy is primarily inductive (Strauss & Corbin, 1990). However, the method also draws on the strengths of positivistic approaches by: (1) providing systematic coding procedures (Strauss & Corbin, 1990) designed to eliminate "specu- lative assumptions not founded on observation," as required by the "empiricist tradi- tion" within positivism (Schweizer, 1998, p. 44), and (2) requiring deductive verification of all findings (concepts and relationships) from the inductive step (Strauss & Corbin, 1990) in a manner that is consistent with the hypo- thetico-deductive logic of positivism (Ackroyd & Hughes, 1992).

The primary objective in this paper is to illustrate the use of, and at the same time, to critically examine the methodological apparatus of grounded theory as described by Strauss and Corbin (1990). To this end, the paper is organized as follows. First, the Strauss and Corbin version of grounded theory methodology (1990) that has gained immense popularity in many arenas of social science since its publication is discussed. Next, the process of coding using an adapted version of Strauss and Corbin's guidelines is detailed, and an abbreviated version of the model of virtual team development that is grounded in data generated during virtual collaboration is presented. Finally, the paper summarizes contributions intended to provide helpful guidance to future researchers utilizing the grounded theory methodology.

A Summary of Strauss and Corbin's Version = of the Grounded Theory Methodology

The grounded theory approach, according to Strauss and Corbin (1990, p. 24), is a "qualitative research method that uses a systematic set of procedures to develop an inductively 3 derived grounded theory about a phenomenon." The methodology is designed to help researchers pro- duce "conceptually dense" theories that consist

The DATA BASE for Advances in Information Systems - Winter 2001 (Vol. 32, No. 1) 39

of relationships among concepts representing "patterns of action and interaction between and among various types of social units" (Strauss & Corbin, 1994, p. 278). Sources of data for developing grounded theory include interviews and field observations, documents, and videotapes (Strauss & Corbin, 1994). A unique aspect of grounded theory is the fact that data collection (or sampling) and data analysis are undertaken simultaneously, and not sequentially as in many traditional methods.

At the heart of the grounded theory methodology, are three coding procedures that Strauss and Corbin (1990) refer to as open coding, axial coding, and selective coding. These codes are generated and validated using the constant comparative method, and coding, at each stage, termi- nates when theoretical saturation is achieved with no further codes or relationships among codes emerging from the data.

Open coding involves "breaking down, examining, comparing, conceptualizing, and categorizing data" (Strauss & Corbin, 1990, p. 61), often, in terms of properties and dimensions. The exami- nation of data in order to fracture it and generate codes could proceed "line by line" (most tedious but most generative, and therefore, often recom- mended in the initial phases of analysis), by sen- tence or paragraph, or by a holistic analysis of an entire document. The open coding process, while procedurally guided, is fundamentally interpretive in nature, and grounded theory researchers "must include the perspectives and voices of the people" whom they study (Strauss & Corbin, 1994, p. 274). Data, for open coding, is selected using a form of theoretical sampling known as "open sampling." Open sampling involves identifying situations/portions of the transcripts that lead to

2 The grounded theory methodology was jointly formulated and first articulated by Glaser and Strauss (1967). The version of the methodology by Strauss and Corbin (1990) is more explicit about the coding procedure, and has become immensely popular in many arenas of social science. However, this version has been severely criticized by Glaser (1992) for misrepresenting/violating the basic tenets of grounded theory described by Glaser and Strauss (1967).

3 Even though grounded theory methodology is primarily involves inductive (ground up) theory building, the methodology also promotes a deductive testing of emerging codes and relationships through the use of the constant comparative method (Glaser and Strauss, 1967; Strauss and Corbin, 1990).

greater understanding of categories and their properties, it is important to note that grounded theory coding and sampling must never be dele- gated to hired assistants, but must be done by the researchers who have a stake in the theory emerging from the project (Glaser, 1978).

Axial coding refers to the analytic activity for "making connections between a category and its sub-categories" developed during open coding (Strauss & Corbin, 1990, p. 97); i.e., reassembling fractured data by utilizing "a coding paradigm involving conditions, context, action/interactional strategies and consequences" (p. 96). Strauss and Corbin warn researchers that "[u]nless you make use of this [the paradigm] model, your grounded theory will lack the density and preci- sion" (p. 99). During the process of axial coding, the relational and variational sampling technique is used, where data is sought depending on its ability to suggest relationships among a category and its sub-categories, or its ability to support or falsify a plausible relationship of a category with its subcategories.

Selective coding involves the identification of the "core category" (central phenomenon that needs to be theorized about) and linking the different categories to the core category using the paradigm model (consisting of conditions, context, strategies, and consequences). Often, this integration takes the shape of a process model with the linking of action/interactional sequences. In creating a process model, the researcher, according to Strauss and Cobin (1990, p.144):

...must show the evolving nature of events by noting why and how action/interaction - in the form of events, doings, or happenings - will change, stay the same, or regress; why there is pro- gression of events or what enables conti- nuity of a line of action/interaction, in the face of changing conditions, and with what consequences.

The theoretical sampling strategy of discriminate sampling is used to select appropriate data at this stage such that weak connections between the categories can be inductively strengthened, and relationships that have already emerged can be deductively tested.

Another issue that emerges at this stage is the role of theoretical sensitivity. Strauss and Corbin


(1990, p. 46) see theoretical sensitivity as "the ability to recognize what is important in data and to give it meaning" by drawing on the literature and personal experience, and by interacting with the data. In other words, the grounded theory methodology does not view "inductive theory building" as implying that the researchers need to flush out their pre-existing theoretical conceptions or knowledge about the phenomenon under investigation, and just let the data speak for itself. In fact, the background that the researcher brings to the interaction with data often leads to creative and important insights (Strauss and Corbin, 1990, pp. 43-44). However, the grounded theory researchers are required to be self reflective so as to be wary of potential biases resulting from their backgrounds, and rigorously validate categories and hypotheses emerging as a result of their theoretical sensitivity.

Methodo logy - An Adapted Grounded Theory Approach

In this section, the study, the data, the coding process, sources of theoretical sensitivity, and deviations from Strauss and Corbin's grounded theory methodology (1990) (along with the justifi- cations) are presented.

An Overview of the Study and Data Sources

Virtual teams were comprised of students from a Canadian university (UA) and a US university (UB), working collaboratively to study a business information systems problem, converting it into a sys-

terns design, and then developing a working prototype. Each virtual team consisted of 4-5 "inter- nal" or "local" group members who were matched with 4-5 "external" or "remote" members from the other university. Thus, each virtual team consisted of about 8-10 members drawn almost equally from UA and UB. The projects lasted for about 14 weeks. A total of 12 teams participated in this study m five in the Fall of 1997, and seven in the Spring of 1998. Table 1 lists the participants in this project. Table 2 summarizes the different events of the project with the associated time lines.

A number of communication and coordination technologies were available to the virtual team- members, including the Webboard, e-mail, videoconferencing, faxes and telephones. Webboard was officially designated as the primary channel of communication throughout the life of the project. While e-mail allowed communication on a one-to-one basis, the Webboard allowed communication to take place in a public domain, and thus, be visible to all other students as well as the faculty involved. While all communication on the Webboard was automatically recorded and saved, students were asked to provide all e-mails exchanged, and post "minutes" of all videocon- ferences, telephone conversations, and Internet chat sessions. While in virtual teams, the members can be spread across multiple locations in time and space, participants in this study were located in two geographically separate locations, representing a dyad. The two sides of the dyad are referred to as UA and UB.

Main Participants Brief description

UA members Members of virtual teams who were students at UA. Primarily involved in interacting with the clients and defining information and end-user interface requirements

UB members Members of virtual teams who were students at UB. Primarily involved in logical design and implementation of the system based on specifications created by UA members in their teams.

PA Professor facilitating the virtual teams from the UA side.

PB Professor facilitating the virtual teams from the UB side.

Companies (each team Located in the same city as UA. UA members interacted interacted with a different with company representatives to define the systems company) requirements.

Table 1. Project Participants


Phase of the Project Timeline Event Descriptions (in weeks)

Phase I: Formation of the team and creation of work plans

Phase I1: Defining the business problem

Phase Ill: System design, development and delivery

Week 0 to week 4

Week 5 to week 8

Week 9 to week 14

• Event 1: Creation of the virtual team by PA and PB • Event 2: Selection of organization by UA members

for which the virtual team would develop a system • Event 3: Introductory videoconference #1 • Event 4: Completion of project proposal by UA members

• Event 5: Completion of the Information Requirements Document (IRD) by UA members

• Event 6: Videoconference #2 to clarify the contents of IRD

• Event 7: Completion of conceptual/logical design by UB members

• Event 8: Completion of user interfaces by UA • Event 9: Prototype delivery by UB members and joint

presentation in videoconference #3

Table 2.

Data Collection

Data was collected from several sources at different points in time. Two main types of data were collected: the communication transactions among virtual team-members (both public and private); and team-members' reflection of their experience at the end of the project. Table 3 summarizes the data collection efforts.

Data Analysis

Formal Project Structure

laboration. Researchers also identified additional readings in many of these areas such as symbolic interactionism, while "guarding against becom- ing captive" to any literature (Strauss & Corbin, 1990, p. 56), and paying heed to Glaser (1978, p. 31), who states

It is vital to read, but in a substantive field different from the research. This maxi- mizes the avoidance of pre-empting, pre- conceived concepts which may easily detract from the input...

Data analysis was done in both informal and formal sessions. Informal data analysis started as soon as the virtual teams were "formed" in the first week of the project. The three researchers (two of them played the roles of participant observers) informally interacted with the data as messages were posted and events started unfolding, deriving theoretical sensitivity from the process of interaction with the data and among themselves, consistent with the recommenda- tions of Strauss and Corbin (1990). At this stage, the researchers also started identifying aspects of their own backgrounds - familiarity with social theories such as actor-network theory, structuration theory, social construction of technology, and symbolic interactionism, different communication and coordination technologies, methodologies such as grounded theory, case studies, hermeneutics, and action research - that could be brought to bear in the theorizing about virtual col-

Being located in two different universities in different countries, the researchers were collaborating in a Virtual environment, and this personal experience created further sensitivity to virtual team related issues, as suggested by Strauss and Corbin (1990).

Coding

Formal data analysis started after the data collection had been completed. Concepts of hermeneutic processes of distantiation, autonomization, social construction, and appropriation (Lee, 1994), suggested that it would be difficult (though not impossible) to generate codes that would be true to the realist ontological assumptions underlying the open coding procedures, if the researchers were separated in space and time during coding. As a result, coding was done during three research trips when all three researchers could work together, face-to-face. It is useful to note that the familiar notion of inter-rater reliability was


Source of data

Webboard

Electronic mail

Videoconferencing

Participant/direct observation by facilitators

Final team reports on project

Reflection documents

Evaluations of other team- members

Nature of data collected

All messages and attachments posted

• E-mails sent directly to facilitators, PA or PB

• E-mails exchanged among team-members

• Real-time observations by facilitators during the meeting

• Videotapes of the meetings

Informal feedback from participants and direct observations

Substantive description of the problem, the design, development and collaboration process

Summary of individual experiences in the project, and lessons learned

Quantitative and qualitative feedback on team-members' performances through e-mail to the facilitators

Time/frequency of data collection

Throughout the life of the project

• At different times • Compiled and submitted by

each team at the end of the project

During the three videoconferencing sessions

Throughout the life of the project

At the end of the project



On-line feedback (optional) Comments on the virtual As and when completed by team project itself participants

Table 3. Data Collection Summary

not applicable to the codes; rather, the three researchers worked together and came to an agreement with respect to most codes 4.

Open coding was done line-by-line initially, and thereafter, open sampling and open coding was done at the webboard message/e-mail message level. Other documents such as reports, reflec- tions, diaries, etc. were sampled and coded at a document level. Well over 200 codes were generated using the open coding procedure. Also, during the open coding process, as predicted by Strauss and Corbin (1990), a portion of axial cod-

4 In the occasional case of lack of agreement regarding the applicability/relevance of a certain code, the researchers opted to be inclusive regarding each other's proposed codes, and relied on the "hermeneutic circle" to validate or invalidate the codes (as each one encountered more data or saw the same data in different light).

ing was done informally (linking sub-categories to categories), as codes were generated and refined.

Table 4 illustrates the open coding process. These are the first three messages posted by the UA members of a particular team. The UB members had not posted any message up to this point. The only other messages that had been posted earlier, were from the facilitators, who listed the members of each team, and provided some guidelines for communication and collaboration.

A few points regarding the open coding process are worth highlighting. First, coding is hermeneutic - that is, coding is an interpretive act of the researchers who are sensitized to certain theoretical concepts (though they need to be committed to making their biases explicit to the extent possible). Also, the codes emerged and continued to


Message

Hi there in UB, rm Henry. I just wanted to say hello and provide you with the rest of our group members' e-mail address. [Names and e-mail addresses] Well, I guess we'll see each other on Saturday at the videoconference

Hello UB, Just letting you know that you are free to e-mail us anytime. I might be

Post date, Week#, Time

1/22/98, week 1, 1:41:52 PM

1/26/98, 1. week 1, 2:56:37 PM 2.

getting an ICQ account going so 3. that if any of you are into real-time 4. chat and wish to communicate that way, it might be something to try 5.

This is Johnny here from the UA. I think that it would be nice to start collaborating on a regular basis so that we can get into the groove of doing it regularly. We will e-mail you as well as post information on the webboard so please keep updated by checking both regularly. Thanks and I look forward to working with you on this project.

Sample Codes generated (underlined) & notes

1. Leadershio - initiative to represent 2. Establishing team's co-presence on the internet 3. Implying preference for communication

technolo.qy (e-mail) 4. Implying technoloav (VC) can bridge the

time and space gap.

.

1/28/98, 1. week 1, 2:44:06 PM 2.

3.

4.

UB members' identitv viewed at an level (as in msg. #1) Collaosina/bridaina across time boundaries Invitation Implying preference for communication technoloav Properties of communication technology/ medium (real-time, synchronous?) Novelty of technolo.qy, recognizing the need to try/exDIore

Self-identity associated with IocaVphysical affiliation. Proposin.q norms of collaboration, technolo.qy-use, and timeliness Implying preference with respect to technology; only partially acceptin,q and modifying the external norm of technology (webboard use) suggested byprofessors. Expressing focus on joint future

Table 4. Illustration

be refined over the life of the project in a process that may be described as the "hermeneutic circle" (Lee, 1991).

Second, a significant majority of codes that were ultimately "used" in the theory had been recurrent. For example, Table 3 shows the open codes from the first three messages of a team, and several recurrent themes such as preference for a particular technology, time gaps/boundaries, and identity are already present. Finally, not all codes were related to individual lines or messages. For example, two important codes emerged while considering the three messages as a unit or "stripS, " and comparing it with comparable strips, i.e., messages posted in other teams during the same stage. One was the directionality of the messages. It is clear that in this case, all rues-

of Open Coding

sages were being posted by UA members, with no acknowledgement or response by UB members. In fact, it was not even clear if UB members were co-present, i.e., monitoring the communication media linking the virtual members. Another interesting code that emerged was the nature of the messages (social, project-oriented, mixed). It is clear that UA members in this particular team did not feel it necessary to develop a rapport with remote members at UB, by talking about weather, games, music, personal interests, before focusing on project related norms. Some of the other

5 According to Agar (1986, p. 28) "a strip is a bounded phenomenon against which an ethnographer tests his or her understanding." We adapt this concept to signify a "unit" of datum that we use to generate or verify a concept or a relationship.


teams, for example, talked about the Lewinsky affair with Clinton, Olympic hockey, drinking, rock groups, etc., presumably to build a social relationship with virtual partners.

Through the use of the constant comparative method, labels of categories continued to be merged, changed, and occasionally, eliminated. As the researchers started feeling comfortable with the emerging categories/sub-categories, they entered them into the NUD*IST database and linked the categories/sub-categories with some of the associated strips of data within the transcripts that had already been imported into NUD*IST (Richards and Richards, 1994; Gahan and Hannibal, 1988). It should be emphasized here that NUD*IST was not used to code the data, but rather just to record the codes, and link them with the "raw" data that had contributed to the codes' emergence. NUD*IST allows the representation of the rich many-to-many relationships that existed among codes and the strips of data.

However, the researchers were unsuccessful in developing dimensionalized properties of each category/sub-category, proposed by Strauss and Corbin (1990). The problem during coding was that it was difficult to distinguish between properties and sub-categories in many instances. Furthermore it became clear, given the large number of categories/sub-categories that were emerging, that it would be virtually impossible to hypoth- esize and deductively validate relationships (among sub-categories and categories during axial coding, and between the core category and other categories during selective coding) based on all combinations of properties 8. After struggling for several days with what was a cumbersome procedure without commensurate benefits, the researchers decided to depart from this specific methodological guideline on open coding.

The goal of axial coding is to facilitate the linking of sub-categories with their respective categories. Strauss and Corbin (1990, p. 99) propose a "paradigm model" (see below for a simplified version) for structuring this linking process:

6 Interestingly, we did not find this methodological guideline in other articulations of the methodology in the literature, or find evidence of adherence to this guideline application in grounded theory research (e.g., Glaser, 1992; Orlikowski, 1993; Urquhart, 1997).

Causal Conditions --) Phenomenon --) Context - ) Intervening Conditions - ) Action/Interaction Strategies -) Consequences.

While the concepts embedded in the "paradigm model" were helpful in thinking about ways to relate sub-categories with categories, the structure of the paradigm model was too mechanistic and thus constraining. Not all sub-categories surrounding a category could be neatly categorized as causal conditions, context, intervening conditions, action-interaction strategies, or consequences. In addition, in many cases, determinis- tic relationships between the causal conditions and phenomenon, and the phenomenon with consequences, were not apparent from the data, and needed to be forced. In trying to fit the data to the paradigm model, the researchers began to understand the severe criticism of the Strauss and Corbin version by the co-originator of the grounded theory methodology (Glaser, 1992, p. 123):

When I first read Basics of Qualitative Research, I was outraged at the non- scholarly changes in grounded theory, the reversion to the verificational approach and the required paradigm, the putting back of all the ills .of preconception and forcing into the method, why it didn't appear like what he (Strauss) had truly read and understood what we had written together, why he seemed to ignore my fundamental inputs into his version of grounded theory and so forth (emphasis added).

Instead, the researchers attempted to accomplish the objective of "axial coding" (i.e., connecting sub-categories with categories) using this 2-step process:

(1)The major categories (e.g., technology, norms, social practices, stages of team development, frames of reference), were hier- archically related to sub-categories, and entered into NUD*IST (Gahan & Hannibal, 1998). For example, the category of technology was linked to the subcategories such as purpose of technology, nature of ownership, accessibility (by time, location, cost, aware- ness), future potential, degree of novels and interconnectedness. At the next level, purpose of technology was linked to information sharing, triggering effect, relationship man-


I Class

I Ownership

I Accesibility

- Novelty

I I I

Purpose

I i i I

II I! It 'n'en''t'll e or' 'Brn .eatons coaton e I I II Information Triggering k epmg/ Justification time and management Defining

sharing t e r r t o ~ Team memory space

Media [ Characteristics I

Future potential

Inter- J connectedness I

Figure 1. Axial Coding

agement, self and group identity articulation as territory definition, record keeping or team- memory functions, justification, and bridging time and space (see Figure 1). Similarly, other sub-categories were linked to other sub-sub- categories. Note that, during this step, the researchers had to revisit and refine the open codes, thereby alternating between the "dis- tinct analytic procedures" of open coding and axial coding modes (Strauss & Corbin, 1990, p. 98).

(2) Researchers created an integrative memo on each of the major categories that was interpretive in nature, attempted to integrate as many sub-categories as possible within the memo on a category. As in the case of categories/subcategories identified through open coding, the memos (the outcome of axial coding) also continued to evolve during sampling for axial as well as selective coding. Presented below is the first draft of an integrative memo 7 (with minor modifications to improve readability) for the category "technology," as entered into NUD*IST :

for a Category - Technology

Collaboration across time and space requires mediation by technology for both symbolic and substantive purposes. Substantive purposes include sharing information, record-keeping, man- aging relationships, pacing, and triggering of activities in collaboration. Some symbolic uses of technology involve the articulation of self and group identity and legitimizing different courses of action by appealing to the use of technology.

Different classes of technology provide different capabilities, some of them different from the fea- tures of technology as defined from the designers' or the implementers" points of view. For example, we wanted Webboard to be a public record.., stu-

i

i dents have extended this use by creating a local enclave for information exchange with local members in a domain traditionally thought of as being public. The Webboard has also become a project

7 Notice that this memo attempts to integrate the sub-categories and the sub-sub-categories of technology presented during the discussion on open coding. Also, we do not claim that this memo is exemplary in the way it integrates the sub- categories of technology. It is provided to merely illustrate how the objectives of axial coding may be achieved to a great extent through the use of this memo.


• • ~i~iii~ ~Ji:y~ii~:

HOW to develop a story ii ii;iiii!iYliiiii;i'Y

line from the categories? Through the use of

Figure 2. Categories that Needed to Be Linked through Selective Coding

archive, conserving team memory through the documentation of agendas, minutes, project steps, and de/iverables.

Virtual teams build their own sense of interconnectedness into their practices., switching between e-mail, Webboard and fax. Threads of communication appear to be unique across teams.., it depends on personalities involved, task at hand, accessibility to technology, preferences (more individual than group), immediacy of tasks, time deadline.

Videoconference was seen as positive not because of project experiences but because it was viewed as highly novel with future career

i potentia/. In some cases, videoconferencing was seen as the project/

Multiple tasks and technologies were used in the project, they were effective for some aspects of the project but not others. Depends on how people have appropriated a communication technology for different tasks. The two popular theories are information richness and social definition 8. Our experience is that neither seems to explain the situation...

In understanding the local appropriation of technology, social definition is only one part of the equation. Manner in which social definition takes place because of the virtual group setting needs to be investigated 9.

8 An example of theoretical sensitivity. 9 This, for example, became a criterion for future theoretical sampling (relational and variational sampling).

Use of different technologies dependent on the interconnectedness with other technologies and people at a particular time and location influenced the use of different media (technologies) for different purposes... All kinds of problems involved with booking locations and synchronizing for videoconferencing. While it has broader "band- width," it has other contextual constraints. Participants (students) tended to see it negatively as videoconferencing ruined their Saturdays. Perceptions regarding technology get shaped by the context, which influence the participants' technological frames and their use of technology.

As mentioned earlier, similar integrative memos were developed for other major categories as part of axial coding. The major categories at this stage, in addition to technology, were communication, norms, social practices, frames of reference, and stages of team-development (see Figure 2).

The objective of selective coding is to explicate a story by identifying a core category and linking the other categories around the core category. Identifying and committing to a story line is no easy task since, as Strauss and Corbin (1990, p. 119) themselves observe, "one is so steeped in the data that everything seems important, or more than a single phenomenon seems salient." Strauss and Corbin again propose the "paradigm model" as the solution, arguing that the model would "grammatically" facilitate the arrangement, rearrangement and linking of categories. While agreeing with Strauss and Corbin regarding the


need of a framework to identify and develop the story line, the researchers feel that the one best approach of forcing the same "paradigm model" in every grounded theory project, is limiting and not true to the idea of "emergence," as Glaser (1992) has argued. Instead, the researchers feel that broad theoretical frameworks (meta-theories) l° that the researchers identify as relevant based on their interaction with the data may be more useful. In this research the meta-theories of symbolic interactionism (Couch, 1996; 1989) and structuration (Giddens, 1976; 1979; 1984) provided the mechanism for identifying the "core category" from the six categories (technology, norms, social practices, frames of reference, communication, and stages of team development) and for explicating a story line that was grounded in the data. The next section provides an overview of the meta-theoretic perspective used to undertake selective coding.

Meta-Theory Used in Developing the Story Line

Upon interacting with the data, the researchers realized that all action, in the context of virtual teamwork where members were not co-located in space and time, had to be understood on the basis of communication. Thus, a framework was needed that would guide understanding of the action that was implied in the communication transactions among members of a virtual team. The work of Carl Couch (1996; 1989) was very useful in this regard. Couch (1996) describes the study of communicative action and practices as key to understanding human societies, and main- tains that any adequate theory of human conduct must use the concepts of communicative transactions, information symbols, and the directionality of communication as the basic units of analysis. These three units of analysis were used as the analytical framewok to develop a micro-level understanding of communicative action in virtual teams. These concepts are now discussed as they appear to apply to virtual teamwork.

These three concepts have to be understood in the context of an understanding of the nature of co-presence in a virtual team. Co-presence refers to the situation in which potential collaborators share consciousness of each-other's presence

10 It is important for researchers to stay away from theories on/related to the substantive area of research.

through some combination of text, auditory, and visual contact (Couch 1989). This expanded notion of co-presence, which goes beyond that of physical situatedness in time-space, provides a richer conceptual lens to study the nature of communicative action based on the three concepts.

Communicative transactions represent the basic units of communication such as messages that are exchanged between the members, using e- mail, computer conferencing, or other appropriate media/technologies. The message contents and the media used, taken in conjunction, has been conceptualized as the communicative transac- tion.

Couch (1996) describes information symbols to be of two types: "referential symbols" and "evocative symbols." Referential symbols denote objects, events, and sequences in the communication process, and are useful in coordinating communication processes. In comparison, evocative symbols help members to develop social solidarity, for example, through the sharing of humor, stories, or personal anecdotes that are not formally (or directly) related to the project.

Communication directionality, in the context of virtual teams with its members distributed across two geographical locations, is of three types. Unidirectional communication refers to the communication pattern where members at one physical location are sending messages but members on the other side are not responding. Bi-directional conversation is said to occur when members from both sides are sending messages, but not actually responding to messages from the other side in a substantive and cooperative manner. In a sense, they are speaking past, or in Couch's terms, with respect to each other since the actors are not sharing affect or co-orientation. Mutual communication develops with growth of shared co-orientation. Members on both sides start to speak with each other, responding to each other's messages, mutually sharing common experiences, and planning/acting not only as per individual or local interests but by taking the priorities, constraints and interests of team-members in other locations.

The meta-theoretical framework derived from Couch's work can thus be summarized with the following four key points (1) parties involved in


communication must be co-present on shared communication channels; (2) communicative transactions are key to develop interpretations of the kind of symbols in use - - referential or evocative; (3) communication transactions also allow us to interpret the directionality of the communication when viewed in conjunction with earlier messages; and (4) by examining the communicative transactions, information symbols, and communication directionality within the specific context of a virtual team project, we develop inferences about the nature of communicative actions.

Giddens' structuration theory (1979, 1984) appeared to provide an avenue to link the communicative action with the structures associated with the virtual teams. (Since structuration theory has been a popular meta-theory among IS researchers, it is not discussed in detail in this paper.) Consistent with Giddens (1979, 1984), structure is not viewed as something physical but rather as a concept that represents memory traces in the minds of the agents engaged in communicative action. Structure is only manifested in the structural properties of social systems as the norms and resources drawn upon by the actors in their everyday action, and interpretive schemes that actors use to make sense of their actions (Giddens, 1979; Orlikowski & Robey, 1991). In this work, structure represents the agents' interpretations of the norms, resources and the shared frames that are embedded in these virtual teams, which are drawn upon to facilitate, legitimize, and make sense of their communicative action.

It is important to emphasize that structure is not static but continuously changing. Over the period of a project, different structures are created, reproduced, and transformed through communicative action of the virtual team-members. These very structures provide a framework within which subsequent action can proceed. This work engaged the researchers in a double hermeneutic process of interpreting the communicative action implied in the open codes of communicative transactions, and the recursive relationship between communicative action and structure (i.e., the stage of virtual team development). The task, during the process of selective coding, was to utilize the meta-theoretical framework derived from symbolic interactionism and structuration (see Figure 3) to string together a theory involving all major categories/sub-categories, while continu-

ously seeking to verify/falsify aspects of the emerging theory using data obtained through discriminate sampling.

Strauss and Corbin (1990, p. 121) argue that it is essential that the grounded theory researchers choose one category as the core category "in order to achieve tight integration and dense development of categories required of a grounded theory." They add, that other important categories may be treated as the core category in a different paper where a different grounded theory may be formulated and articulated. In this work, the theoretical sensitivity derived from Couch as well as Giddens led to the selection of "stages of team development" as the core category (see Figure 4), and to link it with other categories, most notably, communication (and the implied action), norms, and technology. Next, the researchers attempted to link the core category with other categories as well as sub-categories. To maintain the focus in our story line required refining and also dropping a number of sub-categories that had earlier been included in our axial coding memos.

A Grounded Theory of Virtual Team Development 1' (in the Context of a Collaborative Project)

The "story 12'' describes how virtual teams pass through different stages of development during the course of the project. These stages are termed initiation, exploration, integration and completion? 3 It may be worth mentioning here that, even though the labels for the four stages may seem similar to those discussed in the existing group development literature, the specific patterns associated with each stage discussed below are different in many respects to the accepted patterns of communication or action. The researchers believe that these new patterns could not have been discerned without taking an inductive and intensive view of research, as

11 Since this paper focuses on the methodology of the project, only a preliminary version of the theory is presented. A more elaborate version of our theory will be reported in a future paper (currently under review). 12 The word "story" is used deliberately to describe this process theory, in order to highlight the predominantly interpretive methodological approach. 13 It is important to note that different groups have varying rhythms and mechanisms by which transitions take place (or not) from one stage to another.


~ c l n o n t c STRUCTURE his study, "stages of team development" was eptualized as dynamically changing s t r u c ~

Norms, Interpretive Schemes, Resources

Structuration Theory (Giddens, 1984; 1979; 1976)

/ ACTION Check for Co-presence & Examine Communicative Transactions.-, Discemlnforrnation Symbols and Infer Communication Directionality--), Infer Communicative

Iowa Schools' theory of Symbolic Interactionism (Couch, 1996; 1992; 1989)

Figure 3. Meta-Theoretical Scheme for the Study

.... ~::::iiiiiiiiiililliN~i

. . . . . . . . . ~ . . : : i ~ i ~ : : ~ i

Figure 4. Identification of the "Core Category" (Stages of Team Development) Based on Theoretical Sensitivity Based on Figure 3

enabled by the grounded theory methodology. What follows is a brief discussion of the four stages.

team members in different locations have varying backgrounds and experiences with different technologies.

In the initiation stage, after teams have been "formed," distributed members of virtual teams experience considerable ambiguity regarding their roles, the shared goals of the overall project, and the norms by which teamwork should proceed. Adding to this ambiguity and complexity, is the absence of shared norms on the use of communication/ coordination technologies, since

Within these circumstances, the facilitators play a key function in partially defining the context around which the project should proceed. However, the "external norms" proposed by the facilitators have to be appropriated by the virtual team members in order to guide their communicative action. Team members' prior experience with collaborative work and technology mediates


how external norms are internalized by particular members, reinvented, and then proposed to other team members over the communication medium. Initiation involves planting of the seed of an effective social relationship across time and space. Time-space adds to the uncertainty since there is no name to the face, and members are separated by time-zones, geographical distances, and cul- ture, giving a sense of the unknown on one hand, and the wonder of a new experience, on the other. In this context, it becomes important as to which group starts the process of proposing norms, what norms are proposed, the manner in which it is done, and the interpretive schemas through which members try to internalize it.

The nature of communication in the initiation stage is typically unidirectional. Members of the team taking the lead in communication may entrust themselves with the initial responsibility of establishing their own and their local team members' co-presence on different communication media by posting their names, email addresses and ICQ-Ids, etc. The more proactive members may invite remote members to become co-present on the communication media in different ways. Transition to the next stage of exploration takes place as members of the other side of the virtual team establish their co-presence, start monitoring the different communication channels, and respond to the invitations or coercions of remote members.

The exploration stage of virtual team development is characterized by active co-presence of the remote members, being reflected by the fact that messages are posted and monitored by both sides. Each side is in the process of proposing, however tentatively, their sets of norms and expectations about how the project should proceed. While there is evidence of bi-directional communication, there is a notable absence of mutuality. Members of each side monitor each other's presence in the communication technology medium and provide information to each other, without responding specifically to each other's concerns. There is a clear differentiation between intra- and inter-location interests and norms at this stage, with little evidence of merger of interests taking place between self (local team members) and the others (remote team members). Rather than discussing overall project goals, the team members appear to focus on local goals and

concerns. In this stage, team members explore the process by which the norms of communication and collaboration should develop, and also try to deal with their sense of identities. Influencing this process of exploration, is the manner in which information transfer takes place between the team members. Information, when publicly articulated through the use of technologies such as the Webboard, usually allows for its greater visibility to the rest of the team. If widely shared among team members, information can help form the basis for developing a shared focus and doing work in a coordinated fashion. It is important to note that a widespread diffusion of norms may not take place if the majority of communicative transactions use e-mail directed to individuals, since the lack of visibility of information to all team members may make the global acceptance of the norms more difficult. Some members also experience a tension between their local versus global identities during this stage.

To facilitate the transition of the virtual teams to the next stage of integration, which involves a higher degree of mutuality in communication and consequently coordinated action, virtual teams appear to utilize a number of strategies TM, most notably, the building of social solidarity through evocative symbolism.

The integration stage of virtual team development is characterized by the formation of a shared frame of reference, mutuality in communication, and substantive focus in the communications. Shared frame of reference implies that both local and remote members have a common understanding of their goals, their roles, and the norms guiding their communication/collaboration. This leads to the development and articulation of a shared focus and common team-level identities irrespective of the physical locations of the team- members.

In this stage, mutuality of communication is also evident, as team members show empathy and respect for the suggestions, objectives, and constraints (e.g., schedules) of remote members. In

14 It may be useful, from the methodological point of view, to note here, that each of the apparent mechanisms of transition to the integration stage was verified using discriminate sampling of relevant strips. Interestingly, a mechanism of increased frequency of message posting that was suggested by the data initially, did not survive the deductive tests using data obtained through discriminate sampling.


addition, the communication in this stage of integration involves intensive discussion around substantive issues relating to the completion of the project. Since mutual trust and respect has already been established between team members, it becomes easier for substantive discus- sions around project related issues to take place. It is important to recognize that the creation and maintenance of social solidarity through the use of evocative symbols has an important role in sus- taining the integration among team-members. While social solidarity is not a necessary condition for effective collaboration in a virtual team, it is the "glue" that prevents the team (in the integration stage) from regressing to its earlier stages (initiation or exploration) in the event of misunderstand- ings among remote team-members regarding referential symbols TM.

Transition to the final stage of the virtual team development may involve the following two mechanisms: a sense of anticipation of project completion and impending external deadlines.

The completion stage involves the physical closure of the project by handing over the final project deliverables to the project authorities and the subsequent disbanding of the virtual teams. This sense of nearing the end of the project is expressed by virtual team members in either positive or negative ways, depending on the overall project outcome and the experiences that the members have had during the course of virtual work.

There are two ways in which celebration of the joint achievement can be expressed by team members: first, through communication which expresses joy on the completion of the project satisfying external requirements, but in which emotional involvement is at a minimum, as evident from a formal but cordial nature of parting company with remote members. Second, this celebration may be expressed through communication that reflects the positive shared social experiences of members in working together and in suc- cessfully accomplishing work in a virtual team, with participants expressing their emotional

15 It is important to note here that while successful teams conduct most of the project work within the integration stage, unsuccessful teams never really are in this stage, or they keep oscillating between the integration and exploration stages.

attachment with other team-members and signi- fying the pain of parting company.

Alternatively, teams not experiencing success in their project may express their sense of project closure in two ways. First, there could be a sense of individual relief that the stressful experience of virtual teamwork is over. Often, the remaining pri- ority of the team members may be to minimize the damage of negative outcomes to their professional careers. Second, such teams may reach closure by publicly expressing their misgivings regarding other virtual team members, the technology used, and the entire process of collaboration itself.

Discussion and Conclusion

In this paper, the researchers have argued for the need to study new forms of organizations, such as virtual teams, using a predominantly inductive approach. The grounded theory methodology provides an excellent apparatus for inductive theory building. While elements of the popular version of the methodology by Strauss and Corbin (1990, 1994) appear mechanistic, and implicitly encourage "forcing" rather than "emergence," the methodology can be very useful. It is especially useful when adapted in accordance with (1) the nature of the topic of investigation and the data collected, (2) philosophical and methodological assumptions of the researchers, and (3) the accepted norms of methodological rigor required in the academic discipline within which a grounded theory inquiry is being conducted.

Table 5 outlines the similarities and differences between the grounded theory methodology articulated by Strauss and Corbin and the adapted version of the methodology.

Specifically with respect to open coding, and even without the use of dimensionalized properties, the coding procedure ensured that the researchers intimately knew the data. Having undertaken other forms of qualitative research that involve holistic understanding of the transcripts, the researchers are in a position to say that open coding would contribute positively to any genre of inductive qualitative research by facilitating the immersion of the researcher in the data. Therefore, open coding should be consid- ered as a first step for examining qualitative data, whether using case study, ethnography, or any


Aspect of the methodology 1. Data sources

2. Open coding

3. Axial coding

4. Selective coding

Strauss and Corbin's approach (1990) Interview transcripts, field observations, documents, video, etc.

Involves immersion in the data and generation of concepts with dimensionalized properties using constant comparison; primarily an interpretive step.

Identifying categories and mechanistically linking them with respective sub-categories using the "paradigm model"

Selecting a core category and creating a story line about the core category. This story line links other categories to core category. Selection of the core category and linking it with other categories involve the application of the "paradigm model" at the level of categories.

Our Adapted Approach

In addition to all sources mentioned under Strauss and Corbin, we utilized communication transactions generated by participants dunng virtual collaboration.

Almost same; no dimensionalizing of properties involved. NUD*IST used to enter codes and link the codes to appropriate data strips. (The use of a computer package such as NUD*IST, we feel, can unnecessarily add to the complexity of some projects.)

Identifying categories and sub-categories, and arranging them conceptually in a hierarchy (optionally using NUD*IST). Next, generating integrative memo for each category, which involves linking the category to its sub-categories in an interpretive manner. However, the patterns suggested in the memos are continually challenged and validated.

Same; we argue that the "paradigm model" is one possible framework that may be used. The alternate model that we propose based on our theoretical sensitivity is structurally more flexible and interpretive.

Table 5.

other similar methodology. The researcher should, however, be aware that open coding (and open sampling) could continue endlessly (Flick, 1998) unless the researcher uses his/her instincts to ter- minate the process.

A Comparison Between the Strauss and Corbin Approach and Our Adapted Approach in the Sutdy

this form, they could learn a great deal about the important categories.

In attempting to achieve the goals of axial and selective coding, the researchers were forced to sort the codes, think of possible relationships among them, and refine the codes as well as the relationships among them. However, they did not adopt the "paradigm model" that Strauss and Corbin (1990) proposed. Instead, they used the tactic of axial coding, to create memos for each category, that attempted to link important sub- categories to the category. By viewing the data in

It is extremely difficult to undertake selective coding without the guidance of a (meta-theoretical) framework. The "paradigm model" is one possible framework, with its own set of assumptions and structure of the theory it generates. These researchers used the symbolic interactionist and the structurational perspectives to facilitate the selection of the core category of the grounded theory and then weave the other categories and sub-categories into this theory. Without this theoretical sensitivity acting as a "scaffold" (Walsham, 1995), they would be hard-pressed to identify a core category and to link different categories into a coherent theory.


It is important to point out that, like "dimensionalized properties" and the "paradigm model," the criterion of "theoretical saturation" for terminating coding, while theoretically useful, is of little prac- tical value. In this connection, Flick (1998, p. 187) comments that, "The criterion of theoretical saturation leaves it to the theory developed up to that moment, and thus to the researcher, to make such decisions of selection and ending" making the distinction between method and art very hazy. The position of these researchers is that inductive theory building is ultimately an art, but this does not preclude the use of systematic procedures such as those offered by grounded theory methodologists to ensure a strong structural foundation for the artistry to flourish.

In this paper, in addition to providing a confes- sionalist account of experiences with grounded theory methodology, the researchers have also provided a meta-theoretical framework (see Figure1) drawing on existing social theories that can serve as an alternative to the "paradigm model," especially in the context of virtual collaboration. The adaptation of the coding procedures and formulation of the meta-theoretical framework as an alternate model for selective coding are important methodological contributions. An abbreviated discussion of the theory that has emerged is included as a testimony of the effectiveness of the methodological approach.

In closing, the adapted grounded theory methodology that is described in this paper can accom- modate the creativity and flexibility of interpre- tivism along with the positivist rigors of systematic data collection (sampling), data analysis, and deductive verification of inductively derived codes/relationships. Few methodologies have the ontological and epistemological range of the grounded theory, and other researchers are invit- ed to utilize and evaluate the adapted version of the methodology in their investigations of emer- gent forms of organizations and technologies.

References

Ackroyd, S., and Hughes, J.A. (1992). Data Collection in Context, New York: Longman.

Agar, M. (1986). Speaking of Ethnography, Beverly Hills, CA: Sage.

Charmaz, K. (2000) "Grounded Theory: Objectivist and Constructivist Methods," in Handbook of Qualitative Research, - 2nd edi-

tion, N.K. Denzin, and Y.S. Lincoln (eds.), Thousand Oaks: Sage, pp. 509-535.

Couch, C.J. (1996). Information Technologies and Social Orders, D.R. Maines, and S. Chen (eds.), New York: Aldine de Gruyter.

Couch, C.J. (1992). "Toward a Formal Theory of Social Processes," Symbolic Interaction, Vol. 15, No. 2, pp. 117-134.

Couch, C.J. (1989). Social Processes and Relationships, Dix Hills: General Hall.

Couch, C.J. (1986). "Elementary Forms of Social Activity," Studies in Symbofic Interaction, Supplement 2: The Iowa School (Part A), pp. 113-129.

Flick, U. (1998). An Introduction to Qualitative Research, London: Sage.

Gahan, C., and Hannibal, M. (1998). Doing Qualitative Research Using QSR NUD*IST, London: Sage.

Glaser, B.G. (1992). Basics of Grounded Theory Analysis: Emergence versus Forcing, San Francisco, CA: Sociology Press.

Glaser, B.G. (1978). Theoretical Sensitivity: Advances in the Methodology of Grounded Theory, San Francisco, CA: Sociology Press.

Glaser, B.G., and Strauss, A. (1967). The Discovery of Grounded Theory: Strategies for Qualitiative Research, Chicago, IL: Aldine.

Giddens, A. (1984). The Constitution of Society: Outline of the Theory of Structure, Berkeley, CA: University of California Press.

Giddens, A. (1979). Central Problems in Social Theory: Action, Structure and Contradiction in Social Analysis, Berkeley, CA: University of California Press.

Giddens, A. (1976). New Rules of Sociological Method, New York, NY: Basic Books.

Jarvenpaa, S.L., Knoll, K., and Leidner, D.E. (1998) "Is Anybody Out There? Antecedents of Trust in Global Virtual Teams," Journal of Management Information Systems, Vol. 14, No. 4, pp. 29-64.

Lee, A.S. (1994). "Electronic Mail as a Medium for Rich Communication: An Empirical Investigation Using Hermeneutic Interpreta- tion," MIS Quarterly, Vol. 18, No. 2, pp. 143- 157.

Lee, A.S. (1991). "Integrating Positivist and Interpretivist Approaches to Organizational Research," Organization Science, Vol. 2, No. 4, pp. 342-365.

Orlikowski, W.J. (1993). "CASE Tools as Organi- zational Change: Investigating Incremental and


Radical Changes in Systems Development," MIS Quarterly, Vol. 17, No. 3, pp. 309-340.

Orlikowski, W.J. (1992). "The Duality of Technology: Rethinking the Concept of Technology in Organizations," Organization Science, Vol. 3, No., 3, pp. 398-427.

Orlikowski, W.J., and Robey, D. (1991). "Infor- mation Technology and Structuring of Organizations," Information Systems Research, Vol. 2, No. 2, pp. 143-168.

Richards, T.J., and Richards, L. (1994). "Using Computers in Qualitative Research," in Handbook of Quafitative Research, N.K. Denzin, and Y.S. Lincoln (eds.), Thousand Oaks: Sage, pp. 445-462.

Strauss, A., and Corbin, J. (1994). "Grounded Theory Methodology: An Overview," in Handbook of Qualitative Research, N.K. Denzin, and Y.S. Lincoln (eds.), Thousand Oaks: Sage, pp. 273-285.

Strauss, A., and Corbin, J. (1990). Basics of Qualitative Research, Newbury Park: Sage.

Schweizer, T. (1998). "Epistemology: The Nature and Validation of Anthropological Know- ledge," in Handbook of Methods in Cultural Anthropology, H.R. Bernard (ed.), Walnut Creek, CA: Altamira Press, pp. 39-87.

Townsend, A.M, DeMarie, S.M., and Hendrickson, A.R. (1998). "Virtual Teams: Technology and the Workplace of the Future," Academy of Management Executive, Vol. 12, No. 3, pp. 17-29.

Trauth, E.M. (1997). "Achieving the Research Goal with Qualitative Methods: Lessons Learned along the Way," in Information Systems and Qualitative Research, A.S. Lee, J. Liebenau, and J.I. Gross (eds.), London: Chapman and Hall, pp. 225-245.

Urquhart, C. (1997). "Exploring Analyst-Client Communication: Using Grounded Theory Techniques to Investigate Interaction in Informal Requirements Gathering," in Infor- mation Systems and Qualitative Research, A.S. Lee, J. Liebenau, and J.l. DeGross (eds.), London: Chapman and Hall, pp. 149-181.

Walsham, G. (1995). "Interpretive Case Studies in IS Research: Nature and Method," European Journal of Information Systems, Vol. 4, pp. 74-81.

Warekentin, M.E., Sayeed, L., and Hightower, R. (1997). "Virtual Teams Versus Face-to-Face Teams: An Exploratory Study of a Web-Based

Conference System," Decision Sciences, Vol. 14, No. 4, pp. 29-64.

About the Authors

Suprateek Sarker is currently an assistant professor of information systems at Washington State University. He was formerly an assistant professor at the George Washington University. He received his Bachelor of engineering degree in computer science and engineering from Jadavpur University, India, M.B.A. from Baylor University, M.S. from Arizona State University, and Ph.D. from the University of Cincinnati. His research focuses on the use of qualitative methodologies such as positivist/interpretive case studies, grounded theory, ethnography, conversational analysis, and hermeneutics to study phenomena such as ERP implementation, virtual teamwork, and electronic commerce. His teaching interests include database systems, systems analysis and design, case studies in IS, business data communications, and qualitative research methodologies. E-mail: [email protected]

Francis Lau received his Ph.D. in health informat- ics from the University of Alberta. He specializes in the design, implementation, and evaluation of information technology (IT) in health organizations. He has a diverse background in business administration, computing and medical sciences, with 14 years of professional work experience. He has been actively engaged in IT planning, systems development and implementation, and management consulting during the past 12 years, mostly in health. Currently, he is involved in teaching management information systems (MIS) courses as an assistant professor at the University of Alberta, and conducting action research in the areas of virtual teams, knowledge management, diffusion of IT in organizations, and evaluation of IT effectiveness. E-mail: [email protected]

Sundeep Sahay is an associate professor at the University of Oslo. He received his Ph.D. from Florida International University and was a post- doctoral researcher at the Judge Institute of Management Studies at Cambridge University. A primary theme of his research is concerned with understanding the nature of social implications of information technologies in different cultural con- texts. Taking the sociological perspective, he has


been involved in studying the use and consequences of information technologies in the USA, UK, Canada, India, and Malaysia. A specific research topic has been the introduction of information technologies in the context of "developing countries" focusing on the nature of tensions between the beliefs/assumptions inscribed in technologies developed in the West and the social, cultural, and political value systems of the con- texts to which the technology is being transferred. E-mail: [email protected]


using an adapted abstract - heim.ifi.uio.noheim.ifi.uio.no/~sundeeps/publicationsnew/sundeep... ·...

Documents