GUEST EDITORIAL MBR: RETROSPECT AND PROSPECT

It was hypothesized that paired-comparison judgments of multiple-cue profiles would be a joint function of instruction set,- cue-inconsistency, and individual differences. Thirty-one subjects judged all possible pairs of 22 profiles containing eight cues in an own-control, counterbalanced design. A comparison of instructions providing no attribute for judging similarity with those which did provide an attribute revealed systematic differences with respect to both the number and kind of perceived profile dimensions. However, within each instruction set individual differences emerged with respect to the perceived profile dimensionality which would have been obscured with a group-averaging multidimensional scaling procedure. Further, more individual difference variability was present in the instructional set providing no attribute for judgment. Finally, many subgroups of response-homogeneous judges were found to base their judgments of profile similarity on discrepant cues whereas others did not. The possibility was raised that resolution of inconsistent cues may involve either weighting of one or the other of the cues or utilizing an entirely new cue for judgment.