Use of individually scaled versus normatively scaled predictor cues in policy-capturing research

Abstract

Current policy-capturing models scale the levels of the predictor cues on the basis of normative data collected on a group of subjects. Two studies were conducted to determine whether the performance of these models would be improved by scaling cue values individually for each decision maker. The results of these studies confirmed the hypothesis that the scaling of stimulus cues for each decision maker results in a judgment model that is much more successful in reproducing the decision maker’s responses than models employing the same cue scales for all decision makers. Additionally, it was found that the relative performances of models based on regression weights and those models that employ weights generated by the subject are heavily dependent on methodological variables. It is concluded that if there is to be an understanding of the way people utilize information, not only must there be concern about variable weighting, but there must also be consideration of the subjective experience the individual decision maker has with respect to the levels of each variable.