Analyzing Child Malnutrition in Bangladesh: Generalized Linear Mixed Model Appro

Abstract

Child malnutrition is a serious issue for overall child health and future development. Stunting is a key anthropometric indicator of child
malnutrition. Because of the nature of sampling design used in Bangladesh Demographic Health Survey, 2011, responses obtained from
children under same family might be correlated. Again, children residing in same cluster may also be correlated. To tackle this problem,
generalized linear mixed model (GLMM), instead of usual fixed effect logistic regression model, has been utilized in this paper to find out
potential factors affecting child malnutrition. Model performances have also been compared.