Beyond the Respondent-Interviewer Interaction: Exploring the Performance of a Self-Administered Life-History Calendar

Life-history calendar (LHC) methods have been increasingly used in surveys in the last two decades. There is indeed a general consensus on the fact that the highly structured but flexible approach of the LHC facilitates the memory of past events. Respondent's past experiences provide a context for retrieval of autobiographical memories and they are used as anchoring points and time landmarks for remembering. While there is a general agreement that the LHC methods improve accuracy of retrospective data even for different populations and cultures with a non-linear representation of time, the reasons for this performance (i.e., retrieval processes) are less clear. The literature suggest that the LHC performance is amplified by the interviewer-respondent interaction via conversational and flexible interviewing, either in CATI or CAPI. However, studies using self-administered LHC on youth sex behavior argued that LHC can be used also used in self-administered modes reducing social desirability bias.

In this study we present data from two surveys conducted in Switzerland (the Swiss Household Panel pilot study and the Family tiMes survey) between 2011 and 2013. Results show that self-administered LHC could produce equally valid data in terms of completeness (measured with five indicators: number of residential moves, number of intimate partners, number of children, number of jobs, and number of years of unemployment) than interviewer-respondent interactive interviewing. Moreover, self-administered LHC transmitted by mail data had even higher levels of internal coherence than self-administered LHC completed during face-to-face interviews. We argue that improvements in the layout (i.e., graphical visualization) of the LHC, as well as an increasing popularity of life calendar model in social media, may have increased in recent years the accuracy and feasibility of self-administered LHC for data collection.