Data Quality and Recall Bias in Time-Diary Research: The Effects of Prolonged Recall Periods in Self-Administered Online Time-Use Surveys

Research output: Contribution to journalArticlepeer-review

Abstract

Previous research has shown that a prolonged recall period is associated with lower data quality in time-diary research. In these studies, the recall period is roughly estimated on the basis of the period between the assigned diary day and the agreed collection day. Because this is so rudimentary, little is known about the duration of the mean recall period and its consequences for data quality. Recent advances in online methodology now allow a better investigation of the recall period using time stamps. Using a refined indicator, the authors examine the duration of the recall period, to what extent this duration is related to socioeconomic characteristics, and how a prolonged recall period affects data quality. The authors demonstrate that using online time-diary data collected from 8,535 teachers in Belgium, the mean recall period is less than 24 hr for most respondents, although respondents with many time constraints have extended recall periods. Additionally, a prolonged recall period indeed has negative consequences for data quality. Quality deterioration already arises several hours after an activity has been completed, much sooner than previous research has indicated.
Original languageEnglish
Pages (from-to)115-138
Number of pages24
JournalSociological Methodology
Volume53
Issue number1
Early online date5 Oct 2022
DOIs
Publication statusPublished - 1 Feb 2023

Bibliographical note

Publisher Copyright:
© American Sociological Association 2022.

Copyright:
Copyright 2022 Elsevier B.V., All rights reserved.

Keywords

  • recall period
  • data quality
  • data validity
  • memory issues
  • time-diary methodology

Cite this