The Impact of Inferences on Product Evaluations: Replication and Extension

Abstract

The authors attempted to replicate and extend Huber and McCann's findings about the spontaneous occurrence of inferences between attributes in studies of product evaluations. Those findings provided support for the conclusion that inferential processes can distort the validity of the traditional additive model and results based on commonly used estimation procedures such as conjoint analysis. Using conflicting as well as supportive pairs of product attributes, the authors found no evidence of inferencing between attributes unless inferences were explicitly induced.