An Investigation of Utility-Directed Cutoff Selection

Publisher Website
Google Scholar
Abstract
Although the use of cutoffs in heuristic decision strategies has been widely discussed, few studies have attempted to identify which cutoffs will be used in a particular decision. In this article it is proposed that decision makers choose cutoffs that maximally discriminate retained and rejected alternatives in terms of utility. Thus, cutoff selection is influenced by the utility structures that the decision maker has for salient attributes of the alternatives. A model is tested in which cutoffs are assumed to be positioned where the difference in utility between the two attribute levels that span the cutoff is greatest. Utilities are measured by conjoint analysis and cutoffs are identified by concurrent protocols. The proposed relationships are generally supported; however, the nature of the errors in the model's predictions suggests a modification in which cutoff placement is more conservative. A second model with more conservative cutoffs is shown to work well in specific decisions, but does not generally predict cutoffs better than the original model. In addition, the outcomes from heuristic choice processes were often optimal. Decision makers who use cutoffs may select them rationally, with the expectation of making the best choice as well as simplifying a decision.

This publication has 0 references indexed in Scilit: