Abstract
The purpose of this research is to investigate the processing strategies consumers use to form inferences about missing product information. We evaluate the relative effect of attribute information about a partially described brand and about other fully described brands, the effect of attribute intercorrelations, and the effect of prompting inferences. We find that attribute information about a partially described brand has a greater influence than that about fully described competitive brands, that highly correlated attributes more consistently influence inferences, and that prompting inferences produces substantially different results than less intrusive measures.