The moderating role of decision task goals in attribute weight convergence

Abstract

Prior research indicates that various methods for measuring attribute weights show only weak to moderate agreement in the weight assignments. However, there have been few theoretical bases explaining this divergence, and thus we have little knowledge of the conditions under which the weight assignments converge. The current study shows that one determinant of inter-measurement convergence is the comparability of the underlying processing goals evoked to perform the evaluation task. Specifically, evaluation tasks that evoke the goal of differentiation are susceptible to the prominence effect, which reduces the overall convergence with weights measured by tasks based on individual evaluation goals. We show in two studies that the level of convergence improves significantly when the two measures share the same underlying processing goals. Our results have important managerial implications with respect to selecting an appropriate type of direct weighting measurement in consumer studies to better reflect the actual weight assignments in real decision contexts.