Calendar Instruments in Retrospective Web Surveys

Calendar instruments incorporate aided recall techniques such as temporal landmarks and visual time lines that aim to reduce response error in retrospective surveys. Those calendar instruments have been used extensively in off-line research (e.g., computer-aided telephone interviews, computer assist...

Full description

Saved in:
Bibliographic Details
Published in:Field methods 2015-08, Vol.27 (3), p.265-283
Main Authors: Glasner, Tina, van der Vaart, Wander, Dijkstra, Wil
Format: Article
Language:English
Subjects:
Computer assisted research
Cues
Data quality
Factor Analysis
Feedback (Response)
Internet
Interviews
Online Surveys
Polls & surveys
Preferences
Questionnaires
Regression (Statistics)
Response Rates (Questionnaires)
Statistical Analysis
Visual Aids
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Calendar instruments incorporate aided recall techniques such as temporal landmarks and visual time lines that aim to reduce response error in retrospective surveys. Those calendar instruments have been used extensively in off-line research (e.g., computer-aided telephone interviews, computer assisted personal interviewing, and paper and pen interview) and have been shown to increase the quality of retrospectively collected life course data. The goal of our study was to investigate if calendar recall aids can also improve data quality in web surveys. In a methodological field experiment, we evaluated the effects of adding visual feedback and personal landmarks to our questionnaire with regard to response/break off rates, completeness of retrospective reports, interview duration, and respondent evaluations. The study included 1,451 respondents from a probability-based Internet panel who were randomly assigned to one of the four conditions of the experiment. The results indicate that in the web-based calendar tool, visual feedback properties exerted the most influence on data quality.
ISSN:1525-822X
1552-3969
DOI:10.1177/1525822X14554882