Self-administered event history calendars: a possibility for surveys?

TitreSelf-administered event history calendars: a possibility for surveys?
Type de publicationJournal Article
Year of Publication2018
AuteursMorselli, D, Le Goff, J-M, Gauthier, J-A
JournalContemporary Social Science
Date Published01/2018
Mots-clésevent history calendar, mode comparison, paper-and-pencil, self-adminsitration, survey methods

Event history calendar (EHC) methods have received increasing attention from the life-course surveys that have been used in recent years. According to the literature, the EHC provides high-quality data in retrospective surveys because it replicates the autobiographical memory retrieval processes. EHC interviewing is processed through the visual display of individual life events, phases and transitions on a chronological calendar grid, which allows respondents to effectively link events as well as to identify and correct possible dating errors. Moreover, interactive interviewing facilitates the retrieval mechanism. In this study, we test whether the absence of an interviewer and/or interactive interviewing are associated with a reduction in data quality. This aspect is particularly relevant for surveys, as the absence of the interviewer would allow the implementation of EHC methods in self-administered questionnaires. In Study 1, an experimental design compared the results of self-administered paper-and-pencil EHCs in the presence and absence of an interviewer. In Study 2, a quasi-experimental approach compared the results of an interactive EHC interview with those of a self-administered paper-and-pencil EHC. Neither of these studies showed systematic differences between self-administered and interviewer-administered EHCs. The self-administered mode performs better when the instructions and layout design of the questionnaire are clear and detailed. Our findings suggest that the visual properties of the EHC could be a sufficient condition for collecting good retrospective data in the self-administered mode once the initial burden of the task is overcome.