Study limitation reported by the research below: “The quasi-experimental design with preschool-level (not child-level) assignment means the treatment and control groups had unequal sizes (n=94 vs n=56), and selection bias cannot be fully excluded despite similar baselines.”
