Quasi-experiment-implementation-DoubleDifference_STATA

This is a showcase code from a real quasi-experiment (Diff-in-Diff) evaluation I led with Save the children US in Iraq to estimate the effect of changes in caregiver perception on child rearing practices and early childhood development.

Study objectives

The goal of the objectives is to understand the relationship between changes in caregivers’ practices and children’s early learning and development; Specifically:

• Caregiver participants’ at home practices, positive parenting practices, play-based interactions, self-care and regulation over the course of the intervention
• Male caregivers behavior that fosters non-violent, positive home environment
• Children’s learning and development outcomes measured by CREDI (0-3) and IDELA (3-5)

Data and target group

This study used a longitudinal mixed-methods approach to document changes in perceptions and practices around early child development and caregivers’ practices in communities implementing the ECD Toxic Stress Mitigation Toolkit. The study focused on caregivers and parents of young children, and children aged zero to six years old.

• Iraq Sample

  The baseline sample in Iraq include 495 caregivers and 218 3 to 6 year old children children. At endline, 302 caregivers were matched.

Findings

The study results displayed a significant and positive correlation (effect size) with four caregiver outcome indicators: caregiver child relationship, learning and playing, caregiver resilience, caregiver self-efficacy. On the other hand, the findings were insignificant for the remaining three outcomes: gender-based attitudes, gender-based practices and caregiver perceived stress, suggesting no program effect on these three outcomes. In addition, the heterogenous effect estimates, exploration of effect size by gender, showed that male caregivers had a larger and statistically significant effect size for: caregiver-child relationship, caregiver resilience and caregiver self-efficacy. Female caregivers displayed a higher effect size on only one outcome-- learning and playing activities. Moreover, educational level was a significant predictor of the seven outcomes. family size was, as well, a strong predictor of caregiver perceived stress.