Evaluation of Partially Described Multiattribute Options

The Universi O' of Michigan Thirty-five subjects rated the satisfactoriness of two sets of 125 profiles of hypothetical university courses. The profiles in one set were all complete, containing indicators of the state of each course on each of four attribute dimensions. Fifty of the profiles in the second set were partial representations of profiles from the first set, reporting that the level of one of the attributes was "unknown." The other 75 profiles were identical to the remaining profiles in the first set. Results indicated that subjects tended to devalue the partial profiles relative to the complete profiles (p < .01). Regression equations characterizing evaluations of the 75 complete profiles presented to subjects in both replications were derived. Regression weights for the dimension that was omitted in partial profiles tended to be larger in the replication in which complete profiles were accompanied by partial profiles. The results are discussed in terms of uncertainty avoidance and attention and cognitive strain effects on evaluation policies.