Abstract Social scientists are often interested in evaluating evidence of intergenerational or intra-family transmission of skills, knowledge, attitudes, values and behaviours. Nationally representative household surveys often contain information about both children and their parents over time, hence they are valuable for evaluating sources of variation and commonalities within and across family clusters. Analyses of these data introduce choices regarding how family dyads and clusters are defined. We examine consequences of different approaches to research design, in the context of financial literacy. Using data from the Household, Income and Labour Dynamics in Australia (HILDA) survey, we show that decisions regarding the definition of parent-child clusters can have unintended and important consequences for sample composition, and thereby for how we evaluate sources of commonality and differences in financial literacy; with important consequences for how we understand intergenerational transmission.
Kristoffersen et al. (Tue,) studied this question.