Abstract
This article examines the process by which behavioral frequency judgments are generated in consumer surveys. The results from three experiments indicate that the regularity (periodicity of occurrences) and the similarity (idiosyncrasy of the content of occurrences) of a frequently occurring behavior determine (a) the accessibility of the requisite information in memory, and therefore the process by which the judgment is generated, and (b) the accuracy of the associated frequency report. Further, the use of heuristics such as rates of occurrence result in more accurate frequency judgments than the use of recall-and-count strategies based on episodic recall.