Beyond Average Effects: Incorporating Heterogeneous Treatment Effects Into Family Research

Family researchers are increasingly concerned with causal inference. In this article, I urge family researchers to consider 2 types of causal inference: pretreatment heterogeneity, a consideration of nonrandom selection into a treatment (e.g., divorce), and posttreatment heterogeneity, a consideration of systematic differential responses to a treatment. I detail the heterogeneous treatment effects approach, a method designed to account for both pretreatment heterogeneity and posttreatment heterogeneity. I then review existing research that has implemented this method, paying particular attention to research on family life. Finally, I provide concrete examples of how family researchers can implement heterogeneous treatment effects to answer key research questions.