Long-term associations between early childhood education and academic achievement in low- and middle-income countries

• In Event: Evidence of the benefits of early childhood programs on child development in international contexts

Abstract

Approximately 580 million children under the age of five reside in low- and middle-income countries (LMIC), and over 43% are exposed to stunting or extreme poverty (Black et al., 2017). Recent efforts by governments to address poverty-related problems have been focused on increasing early childhood education (ECE) enrollment. In addition, the past decades have seen a drastic proliferation of low-fee private ECE services across LMIC. Consequently, ECE participation rose globally from 32% in 2000 to 50% in 2017 (UNICEF, 2019). Evidence from the U.S. using experimental and quasi-experimental designs show effects of ECE on achievement lasting until fourth grade (Ladd, Muschkin, & Dodge, 2014). However, evidence is mixed when it comes to the long-term impacts of ECE on children’s development, and a preponderance of this evidence comes from industrialized countries. Furthermore, knowledge concerning ECE-achievement associations across different types of ECE arrangements is limited in LMIC.
This study uses a sample of children from economically disadvantaged communities in three LMIC’s with high ECE availability (Table 1) to assess the long-term associations between ECE and academic achievement. Specifically, this study answers whether there are lasting associations between attendance in ECE and children’s language skills at ages 5, 8, 12, and 15. In addition, it compares the strength of those associations for children who attended public vs. private ECE services. We draw data from the Young Lives Study, a longitudinal multi-method investigation following children in India (N=2,011), Peru (N=2,052), and Vietnam (N=2,000). The data include comprehensive measures of children’s development as well as extensive information about family, school, and community characteristics. Language skills were measured by the Peabody Picture Vocabulary Test, a norm-referenced direct assessment of receptive vocabulary. Multilevel structural equation modeling (MSEM) was used to explore these associations (Heck & Thomas, 2015), allowing the nesting of children within communities. Finally, in order to address selection bias, propensity score weighting techniques were used to test whether significant results from the previous models replicated using this rigorous method of analysis.
Results, presented in Table 2, showed that the persistence of the effects of ECE on language varied by country. Results showed that the effects of ECE on language persist until age 15 in Peru and age 12 in Vietnam, while they faded after age 5 in India. Turning to associations by type, results showed that language skills of children who attended public ECE were comparable to children in private ECE in Vietnam at ages 5 and 8, but were lower at age 12. In Peru, attendance in private ECE was associated with higher language skills in comparison to public ECE across all ages. In India, results without propensity scores showed that children who attended private settings performed better than children who attended public ones. Notably, the inclusion of propensity scores flipped this result, indicating that attending public ECE predicts higher vocabulary at age five. In the future, analyses will be extended to include associations between ECE and math achievement.

Authors

• Laura Betancur, Iowa State University
  Presenting Author

• Portia Miller, University of Pittsburgh
  Non-Presenting Author

• Elizabeth Votruba-Drzal, University of Pittsburgh
  Non-Presenting Author