Abstract
Fit between task requirements, user abilities and system characteristics has both intuitive appeal and empirical support as a driver of performance with information technology. Yet despite the volume of research on the construct, there is no unified theory that encompasses the key elements of the different fit constructs. Different studies employ different definitions of fit, both conceptually and operationally. Furthermore, while greater insight is obtained by considering the dynamics of fit and performance over time, prior work has largely focused on fit as a static point-in-time construct. In this paper a unified theory of fit is developed and a comprehensive fit taxonomy is derived. Finally, the theory and definition are shown to extend to a more dynamic conceptualisation of fit.
Table of Contents
Predicting and explaining how information technology (IT) affects human and organisational performance is a key task for information systems (IS) researchers (e.g. Seddon, 1997; Hitt and Brynjolfsson, 1996; Delone and McLean, 1992). Such research can improve understanding of the business value impacts of information technology (e.g. Davern and Kauffman, 1998), and can yield managerial interventions and design prescriptions for more effective use of IS.
The focus in this study is how IT affects individual task performance. IT value creation becomes concrete and most controllable at the level of the individual user, within a specific task and business process context (Davern and Kauffman, 2000). At this level problems with aggregated economic measures are eliminated; and established theory bases in psychology and the cognitive sciences can be used to predict and explain human behaviour.
Fit between task requirements, user abilities and system characteristics has been shown to be a key predictor of individual performance with information systems. Notable examples include Goodhue’s task-technology fit (TTF) construct (Goodhue and Thompson, 1995) and Vessey’s (1991) cognitive fit construct. Intuitively, a better fit yields a better performance. Beyond this intuitive argument however there seems substantial divergence in the literature as to what actually constitutes ‘fit’. For example, Zigurs and Buckland (1998) present ‘a theory of task/technology fit’ built on Venkatraman’s (1989) work on fit in the strategic management literature. Surprisingly, Zigurs and Buckland do not even cite any of the work related to Goodhue’s TTF construct, or Vessey’s cognitive fit construct.
What is clearly required is a comprehensive theory of fit, from which it is possible to derive a taxonomy of the different types of fit that may drive individual performance with information technology. Without such a theory it is difficult to relate fit to other constructs in the literature. Moreover, without a comprehensive theory of fit the definition of the construct itself is confused. The intuitive appeal of the concept is both a key to its popularity, but also hides the lack of any comprehensive theory and definition.
To date, empirical investigations of fit have largely been static. Little is known of how fit changes over time — how users learn and systems evolve. In part, this is an artifact of the field’s experience with the construct. Research logically starts with a static view because it is simpler. Understanding weaknesses of the static view can enrich subsequent efforts to develop a dynamic theory. Prior fit research has recognised the issue of dynamics, but left it unexplored. For example, Goodhue and Thompson’s (1995) technology-to-performance chain model includes feedback, and Vessey and Galletta (1991) state that cognitive fit is ‘an emergent property’ of their model, although these dynamic aspects receive virtually no empirical attention. A dynamic theory of fit holds the prospect of identifying new interventions for improving user performance with information technologies (e.g. it could provide a basis for determining the sorts of training interventions that may be useful). It is also consistent with trends in the behavioural sciences more broadly, which have begun to focus on the explanation of behaviour as an emergent outcome of individual-environment interactions (e.g. McClamrock, 1995; Port and van Gelder, 1995; Thelen and Smith, 1994; Anderson, 1990).
The purpose of this paper is to present a comprehensive theory of fit that is explicitly able to consider the dynamic aspects of fit. The structure of the paper is as follows. Firstly, a theory and definition of fit is presented. Then, a taxonomy of the different types of fit, followed by an exploration of the dynamics of fit that draws on theory from ecological psychology, is presented. Finally, the conclusions and implications for research and practice are presented.