Abstract
When designing decision support systems (DSSs) to support complex cognitive problem solving tasks, it is important to understand what classes of users (domain novices, intermediates or experts) are likely to benefit from it, and conversely, what types of benefits can one expect a given class of users (such as domain novices or experts) will derive. Unfortunately, this is not yet well understood. In this paper we examine a specific class of DSSs which assist users by generating solution options. We compared the results of this class of DSS for four domains and found that they tend to increase novices’ solution quality more than experts’, and that increase may come at the cost of more total time required to produce solutions (although this is not always the case.) Additionally, in an animal nutrition domain, we found that both experts and novices (particularly experts) tended to insist on detrimentally modifying computer generated solutions even when those solutions started out with very high quality. Lastly we discuss possible explanations and implications for DSS design.
Get full access to this article
View all access options for this article.