BackHome
Refresh: Off
Paper Summary
Share...

Direct link:

Effectiveness Attenuation in Scaled-Up Early Childhood Interventions: Evidence From China REACH

Wed, April 23, 10:50am to 12:20pm MDT (10:50am to 12:20pm MDT), The Colorado Convention Center, Floor: Terrace Level, Bluebird Ballroom Room 3C

Abstract

Research Objectives
This study investigates the sustainability of effectiveness in early childhood development intervention programs during scale-up processes. Since 2017, the China Development Research Foundation (CDRF) has implemented the China REACH project, initially piloting in one township in Bijie, a mountainous region in western China, before expanding to all eight townships in Bijie in 2021. By analyzing the differential intervention effects between small-scale and large-scale implementations, we identify critical considerations for intervention expansion.

Theoretical Framework
We posit that small-scale pilots benefit from concentrated resources and higher degrees of direct involvement, facilitating adherence to established protocols. However, during scale-up, organizational resource constraints may compromise project execution, potentially diminishing intervention efficacy. The China REACH project's reliance on local volunteers for multi-site interventions presents unique challenges, suggesting a possible attenuation of intervention effectiveness during the scaling-up phase.

Methodology
This study leverages CDRF's consecutive project evaluations in 2022 and 2023. The initial evaluation assessed developmental outcomes of children in the small-scale pilot post-intervention while establishing a baseline for large-scale pilot participants. The subsequent evaluation tracked children in the large-scale pilot. We employed the Bayley-IV instrument to measure early childhood development across Cognitive, Language, Motor, Social-Emotional, and Adaptive Behavior domains. For small-scale pilot children, lacking pre-test data, we primarily utilized OLS regression to compare their relative advantages to non-participants. For large-scale pilot children, we applied a Difference-in-Differences (DID) approach to evaluate inter-survey changes.

Data Sources
The study draws on CDRF's 2022 and 2023 evaluations. We selected five townships (one small-scale and four large-scale pilot townships) from the intervention sample for analysis. The initial Bayley-IV assessment included 968 children under 42 months, with 68 from the small-scale pilot township receiving continuous intervention from 6 months of age. The second assessment comprised 483 children, including 159 from large-scale pilot townships who began receiving interventions post-2022. We also collected family background information via parental questionnaires.

Findings
Results indicate that during the small-scale pilot phase, after controlling for family background, beneficiary children demonstrated significantly superior development indicators compared to non-participants, particularly in Cognitive, Social-Emotional, and Adaptive Behaviors, with effect sizes exceeding +0.60SD. Language and Motor skills also exhibited significant advantages, underscoring the initial pilot's remarkable efficacy. DID analysis of children in large-scale pilot townships revealed significant improvements in Social-Emotional and Adaptive Behaviors among beneficiaries in operational townships, with increases of +0.40SD and +0.46SD, respectively. However, no significant progress was observed in Cognitive, Language, and Motor skills.

Implications
This study highlights a trend of diminishing intervention effects during the scale-up of intervention projects, a phenomenon evident in rural child development intervention programs. These findings underscore the necessity of equipping organizations with robust execution and supervision systems when expanding interventions to ensure policy efficacy. It is noteworthy that as the scale-up was in its initial stages during the follow-up survey, intervention effects may improve once operations stabilize. Ongoing longitudinal assessment will be conducted to monitor progress.

Authors