Estimating frequencies of emotions and actions: a web-based diary study

Abstract

Mental health questionnaires often ask respondents to report how frequently they experience different emotions. We report two experiments designed to assess the accuracy of these reports and the strategies used to generate them. Each day for 2 weeks, participants in Experiment 1 filled out a web-based emotions-and-activities checklist. Then, they estimated the diary-period frequency of these emotions and activities and indicated how they generated each estimate. In Experiment 2, participants provided frequency estimates and strategy reports, but did not fill out the checklist. We found that (a) the frequency estimates were quite accurate for emotions and activities, (b) participants relied on memory-based strategies (enumeration and direct retrieval) when estimating activity frequencies, but (c) used self-knowledge strategies (personality beliefs and schematic inferences) somewhat more than memory strategies for emotions and (d) the relationship between strategy use and question type was unaffected by diary keeping. We conclude by considering practical and theoretical implications. Copyright © 2007 John Wiley & Sons, Ltd.