Estimating Heterogeneous Treatment Effects of a Parenting Intervention in the Caribbean

Abstract

In 2015, Baker-Henningham et al. conducted a cluster randomized controlled trial in the Caribbean, integrating a parenting education intervention with routine primary care and measuring its effects on outcomes relating to child development and parental performance. Analysis by multi-level models showed that, overall, the intervention had significant benefits for child cognitive development and parenting knowledge. This paper returns to the data to estimate the intervention’s heterogeneous treatment effects and determine whether certain subpopulations within the trial responded significantly differently from others. Through exploratory subgroup analysis, I find that, for several outcomes, the results of the parenting intervention varied significantly based on the gender of the infant, the age of the mother, and the country of residence.