The Lasting Effects of an Early Childhood Parenting Intervention on Children’s Human Capital in Early Adolescence: A Cluster Randomized Controlled Trial in Rural China

See AEARCTR-0003351. The core component of the original intervention was to establish a parenting center at a central location in each treatment village. The centers provided a structured one-on-one curriculum adapted from Reach Up and Learn, organized parenting activities such as story time and singing, as well as free access to age-appropriate toys and picture books. The intervention was designed to target caregivers of young children in the village and aimed to promote caregiver-child interactions that correspondingly support children’ s cognitive, language, motor, and social-emotional development. The intervention lasted approximately two years for eligible caregiver-child dyads.

2015-12-07

2018-06-08

1. Cognitive skills
Children’s cognitive skills are measured using the Wechsler Intelligence Scale for Children, Fourth Edition (WISC-IV).
2. Reading achievement
Children’s reading achievement is measured using a standardized reading assessment adapted from the Progress in International Reading Literacy Study (PIRLS).
3. Social-emotional skills
Children’s behavioral difficulties and prosocial behavior are measured using the Strengths and Difficulties Questionnaire (SDQ). Their social and emotional skills are also assessed using student assessment items adapted from the OECD Survey on Social and Emotional Skills (SSES).

1. Child outcomes
School adjustment, locus of control, beliefs and expectations, time allocation, and mental health.
2. Caregiver outcomes
Locus of control, beliefs and expectations, time allocation, and mental health.
3. Family environment
The home learning environment, parent-child communication, and the quality of family interactions.

In the current study, we will examine the lasting effects of an early childhood parenting intervention when the children reach early adolescence. Specifically, we will re-contact families from the original RCT (AEARCTR-0003351) for an in-person follow-up measuring children’s cognitive skills, reading achievement, and social-emotional skills, as well as other child, caregiver, and family outcomes. In the original trial, after the baseline surveys for all cohorts had been completed, the 100 study villages were randomly assigned in 2015 by computer within counties to the parenting intervention group (50 villages) or the control group (50 villages). The current follow-up involves no new random assignment, and treatment status remains defined by the original assignment.

Our primary analyses will estimate the intention-to-treat (ITT) effects of assignment to the parenting intervention on children’s cognitive skills, reading achievement, and social-emotional skills. Secondary analyses will examine the ITT effects on additional child outcomes, caregiver outcomes, and the family environment. All ITT effects will be estimated using OLS with county fixed effects and standard errors clustered at the village level. We will also estimate the local average treatment effect (LATE) of program participation by using the original treatment assignment as an instrument for the number of parenting sessions attended, measured from administrative attendance records collected during the intervention.

Not available

In the original trial, the 100 study villages were randomly assigned in the office using a computer in 2015, with randomization stratified by county, to the parenting intervention group (50 villages) or the control group (50 villages). The current follow-up does not involve a new random assignment.

The original study used cluster randomization at the village level, with outcomes measured at the child, caregiver, and family levels.

Yes

We will re-invite each of the families from the original study AEARCTR-0003351. The original trial included 100 village clusters, with 50 treatment villages and 50 control villages.

We will re-invite each of the families from the original study AEARCTR-0003351. The original sample size was 1709 caregiver-child dyads in 100 villages.

In the original study, the sample size by treatment arm was as follows:
• Control: 50 villages, 824 caregiver-child dyads.
• Treatment: 50 villages, 885 caregiver-child dyads.
We expect the final sample sizes to be lower due to survey attrition.