The Effects of Estimation Strategies on the Accuracy of Respondents’ Reports of Cigarette Smoking

Abstract

Many survey questions ask about the frequency of specific behaviors (e.g., consumption of a particular food, use of a product, voting). Until recently, the assumption has been that responses to these items are formed through a process of episodic recall, in which the respondent mentally retrieves all pertinent incidents and then tallies them (e.g., Bradburn, 1983). Survey research focused on potential sources of response error associated with episodic recall (i.e., retrieval failures and telescoping effects).