Information Systems as Wicked Problems

The idea of the ‘wicked problem’ (Churchman, 1967), which advocates a pragmatic oscillation between problem and solution, rather than an attempt to reduce the problem to a series of steps to be followed sequentially, has been particularly helpful to us in conceptualising the relationships between people, organisations and information technology (IT). This conceptualisation was tested in the RAMESES project (Risk Assessment Model: Evaluation Strategy for Existing Systems), using grounded theory (Strauss and Corbin, 1997) as the basis for the methodology. The overall objective of RAMESES is ‘to provide a strategic model for the risk assessment of legacy software systems within SMEs (small-to-medium enterprises) considering business process change.’ Thus the relationship between the organisation, the way its staff carried out its processes, and their legacy IT systems was at the centre of our concerns. This chapter describes how the broad conceptualisation of the problem led to a detailed method to address it and the results available to date.

Introduction

This chapter seeks to cast light on the commonly encountered ‘human-centred’ versus ‘technology’ debate in information systems (IS: Clarke and Lehaney, 1998; Clarke and Lehaney, 1999; Lehaney, Clarke et al., 1999). It takes as its starting point a view which sees information systems as complex, adaptive, human activity systems, enabled by information technology (IT). Two approaches dominate in trying to understand such systems. The first redefines them as purely technical systems, for which a fixed and definable objective can be determined: from this point, the problem becomes one of design. The second approach sees the whole system through the views of the human participants: here, the problem initially is one of debate, aimed at determining a consensus view of the system of concern before moving on to designing relevant solutions. The technical view outlined above might be seen as an attempt to reduce the system’s complexity, by removing the voluntaristic, probabilistic behaviour which the human actors bring to the system. Once this is done, more technologically focused IS managers are on comfortable ground, having redefined the system as one which is highly deterministic, and for which a solution can be achieved through the design of a new or improved system. Similarly, the human-centred view may be seen as excluding technical considerations in order to reach agreement on the part of participants before proceeding further.

Critical Thinking and Human Centered Methods in Information Systems

Over the last four decades, information technology (IT) has permeated almost every aspect of our lives. From its origins in the data processing (DP) departments of large organisations, where bureaucratic operations were automated on mainframe computers, IT has penetrated ever further into all kinds of organisational activity, largely due to the accessibility of the personal computer (PC) in the 1980s and the 1990s. Beyond that, IT is also involved in many aspects of our everyday lives, such as education, leisure and entertainment, now that the boundaries between traditional telecommunications technologies and computer-based systems effectively no longer exist. This permeation has meant that the range of people now closely involved with IT on a regular basis has expanded far beyond the white-coated experts in the early DP departments, with terms such as ‘the information society’ in common parlance. It is therefore hardly surprising that the discipline of information systems (IS) emerged and is now evolving to meet the challenge of analysis and design in this complex and dynamic social environment. Nor is it surprising that IS is moving on from its early emphasis on highly structured formal methods of analysis and design, designed to cope with the machine-like preoccupations of the data processing world, to a far softer, human-centred focus. There is clearly an agenda for improvement for IS; although the notions of success and failure may be problematic in themselves, we hear of IS ‘failures’ which make the evening news at depressingly regular intervals — Y2K, the UK air traffic control project at West Drayton, the recent Passport Agency fiasco, for example. More generally, a wide-ranging (14,000 organisations) survey in the UK carried out by the Economic and Social Research Council and the Department of Trade and Industry (OASIG, 1996) concerning the outcomes of IT investments makes worrying reading, reporting that:

System and Training Design for End-User Error

Errors are a major problem for users. In the distant past, the users of computer technology often were limited to computer professionals with extensive technical training. With the growth of personal computers and the Internet, millions of people without technical backgrounds use computer technology on a daily basis, both at work and for leisure activities. Because errors can be such a problem for the end user, it is important to examine the causes of error, as well as different approaches for assisting the end user. This chapter presents definitions of error, as well as a taxonomy of user error. There are two general approaches for assisting the end user in responding to errors: system design and training design. Both of these are discussed in-depth in this chapter. The purpose of this chapter is to describe the current situation of end-user error and suggest ways to improve the end -user experience.

Excellent Systems Analysts

There is evidence which suggests the software crisis still exists and is negatively impacting both information systems (IS) development and maintenance. Kendall (1992) has reported IS development backlogs averaging 30 work-months. Others (Senn, 1985; Yourdon, 1989) including Kendall (1992) suggest a hidden backlog, users’ plans not even submitted as requests because of the identified backlog, may result in IS development delays of up to four to seven years. Further Laudon and Laudon (1998) have determined that 51 percent of software development projects require up to three times more than the initial budget for both cost and time. The situation regarding IS maintenance is also of concern. Kendall (1992) suggests the IS maintenance software crisis has resulted from problems created in phases prior to programming. This situation is further confounded by the fact that the later in the System Development Life Cycle (SDLC) that an error is discovered, the more it costs to fix (Boehm, 1981).

Topological Transformations

This chapter describes some aspects of the development of VICNET, an assemblage of computers, cables, modems, people, texts, libraries, buildings, dreams and images. It is a system that is difficult to characterise, it is dynamic both in geographical and ontological scope, size and usage. I have attempted to capture some of its nature through the use of several vignettes that may give the reader a small insight into parts of its being, then using some of the techniques and explanatory and exploratory mechanisms available from the field of science studies such as heterogeneous engineering and Actor Network Theory (ANT), I reveal some of the ways that VICNET came into existence. Many computer systems are undergoing continual evolution and it is extremely difficult to discern their configuration and what objects have agency at any given point in time; they can be thought of as open systems as described by Hewitt and de Jong (1984). VICNET, an Internet information provider established in 1994 as a joint venture between the State Library of Victoria and Royal Melbourne Institute of Technology, is one such system; it is being used by a large number of people and public libraries, yet simultaneously it is evolving and being shaped by the technology, the users and the environment of which it is part. Consider the system, VICNET as it is called, as a node of a much larger network. I have attempted to unfold this node to reveal the social and technical worlds contained therein, but I also fold the VICNET node in itself so that it becomes part of a much larger sociotechnical system – the Internet. This process of folding I refer to as a topological transformation and it is by studying transformations of this type that may help us understand how open systems come into being and evolve. In what follows, I provide a brief background to VICNET and the data collection method I used. Next, I discuss some the analytical techniques that are available for those who wish to study the development of technological systems. Following this all-too-brief comment I then present a selection of vignettes that show the varied nature of this socio-technical system. Presenting these then allows me to develop further the idea of social topologies introduced in the section on analytical techniques. In the final section there is some discussion as to why this way of looking at socio-technical systems may be useful.

Bottom-Up Management and System Design

Development techniques almost always use top-down approaches to develop software and business systems. Humans need to simplify the external world by using cognitive models to build a boundary around a problem. These necessary, but artificial, boundaries help us cope with the complexity of the problem at hand. However, this deductive process produces dilemmas, as it leads to misconceptions about the real behavior of systems and the people in them. This chapter will look at system design using the system elements (and their interactions) as the starting point of design, that is an inductive approach. Whilst this will not replace the top- down approach, its use will enhance problem solutions. In a contemporary world of loosely coupled organisational elements, it is necessary to view the system from this perspective to fully understand it. This chapter will offer a preliminary methodology to approach system design using ‘bottom up’ thinking. This view is not the opposite of top-down thinking but a supplement to it. It results in asking questions about the desired system, which are fundamentally different in nature to conventional techniques

The Metaphorical Implications of Data Warehousing

Metaphors have long pervaded the discourse around information technology (IT) design (Johnson, 1994), helping developers to conceptualize technological features and functions, to design human-computer interfaces (Rechtin, 1997; Golovchinsky and Chignell, 1997; Rauch, Leone, and Gillinhan, 1997), and to articulate application requirements (Boland and Greenburg, 1992; Davidson 1996a). Metaphors also play an important role in conveying what Swanson and Ramiller (1997) call the organizing vision for IT innovations. An organizing vision, which develops through the discourse of a community of technology producers, information systems (IS) professionals, business managers, and other stakeholders, provides an interpretation of the applications of an IT innovation and the rationale for its use. Buzzwords are labels or names that come to be identified with an organizing vision and “may serve as a potent metaphor” (Swanson and Ramiller, 1997, p. 463) for conceptualizing the roles, relationships, control mechanism, and work processes associated with the IT innovation.

A Measuring of Task-Technology Fit for Computer-Mediated Communication

A key determinant in the success of computer-mediated communication systems (CMCS) and group support systems (GSS) is the task they are used for (Huber, 1984; DeSanctis & Gallupe, 1987). Task models and theories exist in the domain of non-mediated groups (e.g., McGrath, 1984; Wood, 1986) but application of these to GSS and CMCS has been spotty and the results equivocal (Zigurs & Buckland, 1998). Although research findings repeatedly suggest that the fit between task and computer-mediated communication technology is important, researchers have not yet been able to comprehensively describe or measure the dimensions of appropriate fit. This chapter describes the development and initial testing of an instrument to measure the perceived effectiveness of CMCS based on task type (hereafter PE measure). The PE measure extends prior research in several ways. First, it operationalizes the four major dimensions of McGrath’s task circumplex (McGrath, 1984; McGrath & Hollingshead, 1994), a model which frequently is used as a conceptual framework for studying GSS and CMCS (Dennis & Gallupe, 1993). Thus, it will be straightforward to integrate findings from studies that use the PE measure into the existing literature. Second, all four task types are incorporated into the PE measure, where prior research has focused primarily on generation tasks and, to a lesser extent, choice tasks. This comprehensive view of the overall task construct should benefit the process of theory-building as well as prediction in practical applications. Third, the PE measure has been tested successfully within heterogeneous task domains, suggesting that the instrument has validity and is relatively robust.

Rethinking Stakeholder Involvement

Current practice in strategic information systems (IS) planning seems to be focused on surfacing an organisation’s vision and goals, exploring the potential offered by information technology (IT), and designing information systems to support the fulfillment of the stated goals using the most appropriate technology available (García, 1993; Currid, 1994; Lewis, 1994; Andreu et al., 1996). Methodologies for IS planning usually involve the training and participation of individual employees—but only in so far as they contribute to furthering the pre-set organisational agenda. These methodologies also tend to assume a ‘standard’ role for IS experts: providing expertise in IT/IS management. Most of the literature and the practice of IT/IS development in organisations seem to be focused upon technical issues (Davies and Wood-Harper, 1989), where computer science experts play an important role (Winograd and Flores, 1987). They are expected to provide knowledge to solve problems.