Representativeness is not helpful in studying heterogeneity of effects across subgroups

Abstract

In response to our commentary that representativeness best be avoided, Stang and Jöckel argue that representativeness is not needed in studying uniform effects, but that it is helpful when studying effects that vary across subgroups.1 In our original papers, we did address this point: ‘If you want to study the extent to which an effect varies by subgroup of a third variable, you need to design the research to examine the effect by subgroups’2 and ‘When studying effects across a range …