Journal of Official Statistics, Vol.20, No.2, 2004. pp. 301–317
The Time-line as a Device to Enhance Recall in Standardized Research Interviews: A Split Ballot Study
Wander van der Vaart
Abstract:A split ballot experiment was performed on a time-line procedure that was designed to increase the accuracy of responses to retrospective questions. The time-line unites several aided recall properties and was applied to help respondents to reconstruct their educational history. The data were collected during two main waves (1987 and 1991) of a longitudinal social survey in the Netherlands (N = 1,257). In both 1987 and 1991 respondents were asked about their educational history from August 1983 on. The agreement between the 1991 reports and 1987 reports about the period 1983-1987 was used as a measure of recall accuracy. It was hypothesized that the time-line would enhance recall accuracy regarding the number of educational courses attended, the starting year of the courses, and the entire set of types of courses attended. Additionally, it was expected that the time-line would be especially helpful if the difficulty of the recall task was high that is to say, in the case of a high frequency or low saliency of the courses followed. The general picture of the results is that the time-line procedure improved data quality in most conditions and never resulted in inferior data quality, supporting the assumption that it indeed may enhance recall.
Keywords:Data collection, recall accuracy, memory cues, life history
Copyright © Statistics Sweden 1996-2018. Open AccessISSN 0282-423XCreated and Maintained by OKS Group