Attribute conflict and preference uncertainty: Effects on judgment time and error

This research investigates preference uncertainty generated as a function of specific alternative characteristics during multiattribute evaluative judgments. We propose that preference uncertainty has at least two behavioral manifestations: longer judgment times and greater response error in expressed preferences. We investigate two hypotheses regarding stimulus-based causes of preference uncertainty. As predicted by our attribute conflict hypothesis, greater within-alternative conflict (discrepancy among the attributes of an evaluative alternative) led to longer judgment times and greater response error. As predicted by our attribute extremity hypothesis, greater attribute extremity (very high or low attribute values) resulted in shorter judgment times and less response error. We also found that judgment times and response errors were strongly positively correlated at the item level, consistent with our assumption that preference uncertainty generated by stimulus characteristics is manifested in judgment time and error. Finally, we found that the item-level preference uncertainty effects proposed here operate in parallel with strategy-level, effort-accuracy tradeoffs observable across participants. These findings are consistent with the RandMAU random multiattribute utility model developed in a companion article by Fischer et al.