“…The data also included demographic and contextual variables reported by children’s main caregivers, which I used as covariates in the analyses. In particular, there was information on children’s age in months and sex, caregivers’ age and education (none, primary, secondary, higher, or nonformal), and household wealth as measured by the first component of polychoric principal component analysis of basic dwelling characteristics (e.g., access to electricity) and household assets (e.g., television), as computed in Cuartas et al (2023). Furthermore, the data included information on children’s exposure to other harsh discipline at home, in particular whether caregivers yelled at the child (yes vs. no), a summed index of the availability of books in the home, comprising five dichotomous yes/no indicators (e.g., storybooks, coloring books; possible range = 0–5, α = .75), a summed index of playthings/toys in the home, comprising nine yes/no dichotomous indicators (e.g., toys from a shop, homemade toys; possible range = 0–9, α = .64), and an index of caregivers’ stimulation practices, measured as a sum index of mothers’, fathers’, and other adult caregivers’ engagement in nine play/learning activities with the child (yes vs. no), including reading, telling stories, singing songs, going outside, playing, naming objects, teaching new things, teaching letters, and playing counting games (α = .79).…”