Evaluating Binary Logistic and Multilevel Logistic Regression Model by Analyzing the Factors Associated with Early Childhood Development in Bangladesh.

Abstract

Objective: This study aims to evaluate the factors influencing Early Childhood Development (ECD) in Bangladesh, highlighting its long-term economic and social benefits and the need for equitable access to developmental resources for all children.

Materials and Methods: Data was drawn from the Bangladesh Multiple Indicator Cluster Survey (MICS). Binary logistic regression and multilevel logistic regression models were applied to examine the associations between ECD outcomes and key socio-demographic, economic, and caregiving variables. Model comparisons were conducted using Akaike’s Information Criterion (AIC).

Results and Discussion: Access to children’s books and store-bought toys was found to be significantly associated with positive outcomes. Additionally, the use of internet and mobile phones showed a protective effect. The multilevel logistic model (AIC = 4845.4) demonstrated a better fit than the binary logistic model (AIC = 4928.5).

Conclusion: Both individual and community-level interventions are essential to enhance early childhood development in Bangladesh. Quality early childhood programs foster cognitive, emotional, and social growth, contributing to improved educational outcomes, stronger family environments, and more resilient communities.

Dhaka Univ. J. Sci. 74(2): 257–262, 2026 (July)