The impact of following rules on the sample composition of household based panel studies

Presenter: Nicole Watson, University of Melbourne, Australia

Author: Nicole Watson

Household based panel studies adopt following rules that identify who is followed over time and interviewed. These following rules expand the sample to include new births and adoptions, but may also include the other parent of these births/adoptions, new immigrants and other household members not already part of the continuing sample. Further, it is also normal practice in most household based panel studies to interview all adults living with a continuing sample member each wave. These following and interviewing rules will change the composition of the sample in ways that may not be anticipated by researchers, particularly those interested in household composition and household dynamics.

We examine what impact various following rules have on the sample composition via a simulation study. Drawing from Australian data, we use household changes from the Household, Income and Labour Dynamics in Australia Survey to add births, deaths, and other household joiners and leavers. Applying these household compositional changes to a sample over time, we examine five sets of following rules, including those adopted by Understanding Society and other major household panel studies. We find the sample grows in household types that frequently change (such as young people in shared accommodation or young couples) and reduce in single person and young family households. This effect is somewhat moderated by attrition and varies by the particular following rules adopted.