Anatomy of Treatment Effects on Early Childhood Learning

• In Event: Empowering China's Youngest Rural Children: Uniting Intervention Research and Policy Advocacy for Lasting Impacts

Abstract

Objectives: Building on Author et al. (2024), this study examines the dimensionality and evolution of skills in a program designed to foster the skills of rural Chinese children. Unlike conventional frameworks, it evaluates both traditional test scores and latent skills through an extended IRT model, analyzing the interrelationships and micro-dynamics of skill development.
Theoretical Framework: The study diverges from the traditional human capital framework (Ben-Porath, 1967), proposing a new model for understanding skill formation. It explores the micro-foundations of skill formation, providing an alternative to conventional investment theories applied to firms.

Data Sources and Methodology: The research focuses on children aged 6 to 42 months participating in an early childhood home visiting intervention based on the Jamaica Reach Up and Learn curriculum. Data is collected weekly, ensuring comparability as all participants follow the same curriculum and are assessed on identical tasks. This design eliminates variability in skill measurement across different ages and individuals. The study also compares two measures used in evaluating learning programs, addressing the impact of untaught items and varying assessment intervals on test scores.

Results: The analysis decomposes treatment effects into several components: (a) Direct exposure to new knowledge; (b) Autogenic growth from maturation and environmental exposure; (c) Dynamic complementarity, where previously acquired knowledge aids new learning; (4) Digestion effects, emphasizing the importance of repetition and consolidation; (5) Cross-fertilization, where acquiring one skill facilitates others; (6) Raw ability as a learning factor The impact of untaught and untested items
The study finds that exposure to new knowledge is the most significant factor, with dynamic complementarity enhancing the productivity of later investments. Addressing learning in slower children could further boost treatment effects.

Significance: The study challenges the concept of "fadeout," where treatment effects diminish after an intervention ends. It shows that treated children’s knowledge appreciates over time and with lesson repetition, contrary to the depreciation suggested by traditional models. Control group children exhibit growth consistent with autogenesis and social influences, indicating the importance of interactions outside the experimental scope.
In conclusion, this research offers a revised understanding of skill formation, moving away from the Ben-Porath model and emphasizing the continuous appreciation of knowledge and the role of dynamic complementarity in learning.

Authors

• James J. Heckman, University of Chicago

• Haihan Tian, University of Chicago

• Zijian Zhang, University of Chicago

• Jin Zhou, City University of Hong Kong