Does the suggestion that respondents recall events chronologically significantly influence the data collected?

Abstract

This paper is about estimating a salience scale for trip reporting. The measurement project began as a way of establishing the affects of methodological changes between 1994 and 1997 in the Canadian Travel Survey. This is a survey that Canada uses to study the travel of its residents. There were several changes in methodology that could be expected to influence how trips are recalled. A key change was that in 1994 the interviewers were encouraged to suggest to the respondents that they report chronologically but in 1996 and 1997 they were told not to use the prompt during interviews. It is likely that other changes resulted in response burden being higher in 1996 and 1997 than in 1994. For people reporting two or more trips, chronological reporting should result in trips being reported in random order. It was found that the order in which trips were reported changed with the change in memory cue. The change may have been exacerbated by some trips not being reported because of response burden. The data were analyzed using paired-comparison. The effect of the methodology change is found to be significant. The implications for future research as well as general implications for survey methodology are reported.

Parent Publication

• Proceedings of the 2000 Northeastern Recreation Research Symposium

Citation

Hill, Andrew; Beaman, Jay; O''Leary, Joseph. 2001. Does the suggestion that respondents recall events chronologically significantly influence the data collected?. In: Kyle, Gerard, comp., ed. 2001. Proceedings of the 2000 Northeastern Recreation Research Symposium. Gen. Tech. Rep. NE-276. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 201-204