Although the application domain has long been acknowledged as playing a significant role in IS problem solving, very little research has been conducted into the effect that it has on performance on IS tasks, and even less theory has been developed. In this paper, we present theory that explains the role of application domain knowledge that is contingent upon the structured nature of the IS task under consideration. We then illustrate the theory on the well-structured problem area of conceptual schema understanding, and on the ill-structured problem area of software maintenance.
This research develops a unifying theory of the role of the application domain in IS problem solving that explains the findings from two experiments that focused on the role of the application domain in IS problem solving and that produced different results.
The theoretical framework that we use to form the structure for our theory is a dual-task problem-solving model based on the theory of cognitive fit. Cognitive fit applies not only to problem solving in each of the contributing domains (application and IS), but also to the interaction between the two. The theory of cognitive fit allows us to distinguish different types of interactions between the tasks in the IS and application domains, when the two types of tasks match and when they do not. Those interactions may be supportive, neutral, or conflicting, depending on the structured nature of the problem area under investigation.
In solving tasks in well-structured problem areas, all of the information needed for problem solution is available in the external problem representation and problem solving can take place with reference to IS domain knowledge alone. In this case, knowledge of the application domain plays a role only in solving problems in which cognitive fit does not exist. Analysis of the well-structured problem-solving area of conceptual schema understanding (Khatri et al., 2006) revealed that knowledge of the application domain aided problem solving only in schema-based problem-solving tasks (fit does not exist), and not in syntactic and semantic comprehension tasks (fit exists). When cognitive fit does not exist, the information required for task solution is not available directly in the conceptual schema and transformations are required.
In solving tasks in ill-structured problem areas, on the other hand, the information needed for problem solution is not available in the external problem representation and application domain knowledge is essential to problem solution. When knowledge of the application domain matches the knowledge required to solve the problem, cognitive fit exists and problem solving is facilitated. However, when knowledge of the application domain does not match that required to solve the problem, dual-task interference, which is manifested in an inverse relationship between knowledge of the software gained during problem solving and performance on the IS task, occurs. Analysis of software maintenance tasks (Shaft and Vessey, 2006) revealed that when knowledge of the application domain matched that required to solve the maintenance task, improved knowledge of the application domain during conduct of the modification task was linked to better problem-solving performance. However, when the two types of knowledge did not match, an inverse relationship between knowledge of the application domain and problem-solving performance resulted.
Our theory has implications for research in both IS and cognitive psychology. From the viewpoint of research in IS, there are two major implications. First, the dual-task problem-solving model presents a new way of viewing IS problem-solving. Its foundation in theory in cognitive psychology provides the opportunity for IS researchers to investigate the role of what is acknowledged to be an important and under-researched area of IS problem solving: the role of the application domain. The dual-task problem-solving model and its theoretical underpinnings therefore open the way for the development of a stream of research on the role of the application domain. It should always be remembered, however, that the research needs to be conducted in the context of the degree of structure in the problem area under investigation.
Second, this research adds to the strength of a growing body of literature that further testifies to the pervasiveness of cognitive fit in problem solving (see Vessey, 2006). There are a number of possible avenues for further investigation. For example, the distributed model of problem-solving suggests other factors in the fit models, such as the nature of the internal and mental representations in each of the domains, may either facilitate or inhibit problem solving in a given set of circumstances, and could be the subject of future research.
From the viewpoint of research in cognitive psychology, the community has focused on ‘people’s ability (or inability) to perform two or more activities concurrently’ (Pashler, 1994). The findings of dual-task interference have been pervasive, and research does not appear to have been undertaken to examine other possible types of interactions (and their underlying mechanisms), although a number of authors have observed that problem solvers have a greater ability to perform two tasks that are compatible, as opposed to incompatible, at the same time, thus reducing the impact of dual-task interference (see, for example, Koch and Prinz, 2002; Whitaker, 1979).
What is specific to the types of tasks we investigated is that they either interact with each other, or have the potential to interact. Therefore, instead of focusing on the mechanisms by which dual-task interference occurs, we focused on the circumstances in which knowledge in each of the contributing domains interacts, and the type of interaction that results. Our contribution to the theory of dual-task problem solving in general, therefore, lies in introducing theory to determine when dual-task problem solving results in synergies between the two types of interacting tasks, when it results in interference, and when there are no effects. The cognitive psychology community could extend the focus of its research to determine the mechanisms by which certain tasks that are conducted simultaneously facilitate, while others inhibit, problem solving.