|
Article Excerpt
Abstract
Requirements determination (RD) during information systems delivery (ISD) is a complex organizational endeavor, involving political, sensemaking, and communicative processes. This research draws on the analytic concept of technology frames of reference to develop a socio-cognitive process model of how frames and shifts in frame salience influence sensemaking during requirements determination. The model provides a theoretical and conceptual perspective that deepens our understanding of requirements processes in organizations and of the socio-cognitive basis of power in ISD. The paper reports on a longitudinal case study, in which four technology frame domains were identified and the influence of frames on project participants' understanding of requirements was traced through eight RD episodes. Repeated shifts in the salience of the business value of IT and IT delivery strategies frames disrupted project participants' understanding of requirements and contributed to a turbulent RD process. Analysis of frames an d framing helped explain how interpretive power was exercised, yet constrained, in this project. Implications for further research and for practice are considered.
Keywords: Requirements determination, frames of reference, social cognition, information systems development
ISRL Categories: FB04, AA041, FB03, EF, FB03
Introduction and Motivation
Identifying and agreeing on the requirements for a new information technology (IT) application are two of the most difficult tasks in the information systems delivery (ISD) process. Requirements determination (RD) is characterized by ongoing sensemaking among stakeholders, and it can be chaotic, nonlinear, and continuous (Banseler and B[empty set]dker 1993; Curtis et al. 1988; Walz et al. 1993). Not only are there multiple stakeholders whose various ways of understanding requirements must be taken into account and reconciled, but their ideas may change, particularly in long-term projects or if business conditions and key stakeholders change. We often hear laments about scope creep, project drift, or requirements becoming a moving target. Given these organizational realities, a reasonable goal might be to attempt to arrive at provisional agreements so that something useful could be built and implemented. However, even such a pragmatic goal can be difficult to achieve.
Researchers have examined a number of social, cognitive, and political processes in ISD that influence RD activities and outcomes. A theoretical approach that focuses on sensemaking processes is helpful for investigating why participants in such activities understand requirements as they do (Dougherty 1992) and why their understanding of requirements may change and shift (El Sawy and Pauchant 1988). Orlikowski and Gash (1994) developed the concept of technological frames of reference as an analytic lens for examining how stakeholders' socio-cognitive interpretations influence their actions related to IT development and use in organizations. This research draws on the technology frames concept to develop a process model of how frames and frame shifts influence sensemaking during requirements determination and applies this model in the analysis of an ISD project. The framing model provides a theoretical and conceptual perspective that deepens our understanding of requirements processes in organizations and of t he socio-cognitive basis of power and influence in ISD.
In the next section, a review of the theoretical foundations for this work in the organizational literature on social cognition and its applications in IT research is presented. From this basis, the framing process model is outlined. The methodology section describes the longitudinal field study that provided the empirical basis for the research. Four technology frames were evident during RD activities in this project. The influence of these frames is traced through eight episodes in the project studied. This narrative examines contextual changes and events that triggered shifts in frame salience and the implications of these shifts for provisional agreements about requirements. Recurring shifts in frame salience resulted in repeated reinterpretations of requirements as the project unfolded. How the framing model and frame shifting patterns further our understanding of RD processes is then considered. The paper concludes with a discussion of the implications for research and practice.
Theoretical Foundations
Researchers have for some time been interested in how individuals cognitively process information and how their information processing affects behavior, decisions, and performance. Underlying this interest is the premise that reality is socially constructed through human beings' interpretations of experience and action and their social negotiation of meaning (Berger and Luckmann 1967; Weick 1979). Social cognitive research shares the fundamental tenet that an individual's knowledge about an informational domain is cognitively structured through experience and interaction. Findings suggest that these socio-cognitive knowledge structures provide templates for problem solving and evaluation, focus attention on information consistent with existing structures while masking inconsistent information, and fill information gaps with information that conforms with existing knowledge structures (Fiske and Taylor 1984; Markus and Zajonc 1985; for a comprehensive review of this literature, see Walsh 1995). Although cognit ion occurs at the individual level, researchers have posited the existence of group-level knowledge structures, that is, of shared knowledge and beliefs that function in a similar way as individual knowledge structures (Fiol 1994; Gioia et al. 1989; Porac et al. 1989; Walsh and Fahey 1986).
Socio-cognitive approaches have many applications in organizational and technology studies. Organizational researchers have been concerned with how individual and group knowledge structures influence interpretation of meaning, action, and organizational outcomes, particularly related to strategic direction setting and organizational change (Bartunek 1984; Bartunek and Moch 1987; Fiol 1994; Kiesler and Sproull 1982; Lord and Foti 1986; Weick and Bougon 1986). Others have addressed socio-cognitive processes underlying product development processes (Dougherty 1992; El Sawy and Pauchant 1988; Walsh et al. 1988). Social construction of technology researchers have investigated how knowledge shared by members of a social group (e.g., scientists, inventors, users, manufacturers) influences their development of artifacts and their understanding of a technology's properties and uses in a social context (Bijker 1995; Bijker et al. 1987).
IT researchers have similarly been concerned with socio-cognitive processes, positing that requirements for IT applications do not exist a priori but are socially constructed through interactions among ISD participants (Bansler and Bodker 1993; Curtis et al. 1988; Dagwell and Weber 1983; Malhotra et al. 1980; Newman and Nobel 1990; Walz et al. 1993). In these interactions, the style of interaction and language use influences what requirements are identified and legitimized (Boland 1978; Boland and Greenberg 1992; Davidson 1999; Mason 1991; Salaway 1987). Of particular concern to researchers is how technical designers' assumptions dominate decisions about technology design and application (Boland 1979; Bostrom and Heinen; 1977; Dagwell and Weber 1983; Ginzberg 1981; Hirschheim 1986; Hirschheim and Klein 1989; Markus and Bjorn-Anderson 1987). Such research complements and extends technology-based or methodological approaches by addressing social, learning, and negotiation processes in ISD.
Drawing from these areas of research, Orlikowski and Gash (1994) articulate a socio-cognitive perspective in IT research focused on the concept of technology frames of reference. (2) They use the term frame to describe knowledge structures derived from a domain of knowledge and experience. Technology frames are "that subset of members' organizational frames that concern the assumptions, expectations, and knowledge they use to understand technology in organizations. This includes not only the nature and role of the technology itself, but the specific conditions, applications, and consequences of that technology in particular contexts" (p. 178). In their study of the adoption of Lotus Notes software, they identified three frames that characterized technologists' and users' understanding and use of Notes: (1) nature of the technology, i.e., understanding of its features and uses; (2) technology strategy, i.e., assumptions about management motivation for implementing technology and criteria for judging success of the IT; and (3) technology-in-use, i.e., expectations about priorities and resources, training approaches, ease-of-use, and policies for security and quality.
The technology frames concept provides a useful analytic lens to investigate how ISD participants come to understand requirements as they do and a starting point to investigate the circumstances that facilitate or constrain their maintaining provisional agreements about requirements. ISO participants draw on their technology frames to make sense of contextual information and to ascertain the implications for project requirements. That is, they interpret their organizational worlds to arrive at an understanding of requirements. In this process of interpretation, frames act as templates for problem solving as well as imprecise, conservative filters for new information. In team-based or group project work, a diversity of group and individual technology frames provide an array of possible socio-cognitive filters for interpreting contextual information and making sense of possible requirements. Orlikowski and Gash show that differences, or incongruence, in the frames of key stakeholder groups are a source of disru ption in ISD. However, others have argued that socio-cognitive differences may facilitate diverse interpretations of information and thus improve group decision making in problem formulation activities (Fiol 1994; Walsh et al. 1988; Walsh 1995).
To investigate turbulence and change in problem-setting activities (such as requirements determination), it can be instructive to consider socio-cognitive effects such as frame shifting (El Sawy and Pauchant 1988). Although frames, once formed, are resistant to change (Walsh 1995), contextual changes can trigger shifts that bring new knowledge to the forefront of sensemaking (Bartunek 1984; Barr 1998; El Sawy and Pauchant 1988; Gioia 1986). El Sawy and Pauchant (1988) found that frame shifts can be abrupt and of short duration; nonetheless, these shifts in frame salience influence how participants make sense of environmental information at that point in time, and the decisions and actions they subsequently take. Changes that trigger a shift in salient technology frames could lead to reinterpretation of information and lead to new understandings of IT requirements.
Frame salience may also shift when participation in a project changes. Walsh et al. (1988) demonstrated that the frames of individuals who are highly influential are more heavily weighted in group decision-making tasks. Requirements determination is often driven by individuals such as executive champions, project leaders, or lead designers (Curtis et al. 1988; Heng et al. 1999; Newman and Sabherwal 1996; Reich and Benbasat 1990; Walz et al. 1993). The technology frames of such individuals can become the dominant filter that shapes how IT requirements are articulated and legitimated among ISD participants. If changes in participant power occur during ISD, as it often does (Robey and Newman 1996), frames may likewise shift toward those of dominant participants, altering the basis on which interpretations are made by other participants. Socio-cognitive analysis can thus complement our understanding of power and influence in ISD (Brown 1998; Markus 1983; Robey et al. 1989; Sillince and Mouakket 1997) by drawing a ttention to the basis of power in stakeholders' interpretations and dominant cognitive influences (Markus and Bjorn-Anderson 1987).
The socio-cognitive process model depicted in Figure 1 integrates the analytic concept of technology frames with these empirical and theoretical insights. Point (1) suggests the function of frames as attention-directing and problem-solving templates. Point (2) illustrates that frames act as interpretive filters favoring cues that are consistent with an existing frame. Point (3) depicts the situation in which change triggers, such as a shift in business strategy or changes in stakeholder participation may lead to shifts in frame salience and, ultimately, new understandings or interpretations of project requirements.
This socio-cognitive model provides a theoretic basis and analytic lens with which to investigate turbulence in requirements determination. The framing model suggests that if ISD participants are to arrive at useful definitions of requirements, their frames must be sensitive to changing conditions that arise. At the same time, frames must be stable enough to buffer distracting cues, so that provisional agreements can be engendered and the ISD process can proceed. Instability in RD, such as participants' inability to maintain provisional agreements about requirements, may indicate frequent frame shifts and a framing process gone awry. The next section describes the research design and methods for an in-depth case study in which this model is used to analyze the influence of frames and frame shifts in RD processes. Examining frame shifts helped explain how organizational change and shifting participation in the project influenced participants' understanding of project requirements and disrupted their provisiona l agreements.
Research Study and Methods
The research study reported in this paper was conducted at Group Health Incorporated (GHI) (a pseudonym), a health care insurance company in the eastern United States that offers a variety of health insurance products and services. At the time of this study, there were approximately 6,000 employees located at headquarters and in sales, customer service, and health care service locations. At one time the dominant insurer in its regional market (with over 70% market share), GHI was frequently characterized as bureaucratic and inefficient. Since the late 1980s, GHI has suffered intense competition, the result of market changes, the shift toward health maintenance organizations (HMOs) in the United States and debates over the cost of health care and government's role in this market. GHI management reacted to these challenges by acquiring one HMO, building a second, enacting cost control and staff reduction measures, and hiring executives to focus on rebuilding market share (at a low of about 35 percent at the tim e of the study).
In the early 1990s, the IS department comprised about 10 percent of all GHI employees. An executive characterized the department as "a large utility," which was mainframe-based with disparate, ineffective transaction processing systems, dependent on vendors and purchased software, and with no coherent communications network in place. During the late 1980's, GHI had undertaken a major IS development initiative to replace out-of-date systems, spending over $100 million on the MIS Initiative project, but producing no workable systems. Along with other financial problems, this IS failure brought governmental scrutiny of GHI's financial situation. As a result, in early 1992, GHI executives decided to outsource the entire IS operations and development functions to a vendor, Information Systems, Inc. (ISI). Except for a small internal staff of business analysts, IS personnel were transferred to SI and projects related to the MIS initiative were scrapped. Over the next two years, transaction processing systems were c onverted to the ISI systems platform; this process was still underway during the field study.
The systems development project examined here, the Business Information System (BIS) project, began in 1992 with a requirements study, shortly after IS outsourcing was undertaken. According to project participants and early requirements documents, the project's purpose was to create a relational database of marketing data extracted from the new platform of subscriber enrollment and sales systems, and to provide a user-friendly interface for end-users to generate business analysis reports. Technical staff in the sales organization, working with systems developers in SI, IS personnel at GHI, and external consultants, conducted requirements determination activities over the course of the 30 months examined here. RD had been underway for more than a year before the research project began in mid-1993; it continued throughout the field study and was still going on when the field study ended in mid-1994.
Research Design
Yin (1988) recommends case study research to investigate contemporary phenomenon within its real-life context, when the boundaries between the phenomenon and the context are...
|
