Understanding childhood malnutrition in Nepal: A hierarchical regression approach
Abstract
This thesis investigates the determinants of childhood malnutrition in Nepal, with a particular emphasis on the importance of district characteristics relative to household and child characteristics. Using a new dataset constructed from child and household data from the 2006 and 2011 Nepal Demographic and Health Survey (DHS) and community characteristics aggregated from the 2004 and 2010 Nepal Living Standards Survey (NLSS), we estimate a variety of hierarchical regression models, which allow us to partition variance in height-for-age Z-scores between the levels of different hierarchical specifications, and then to partition that variance further, between group-level parameters. Our findings suggest that the majority of variance in HAZ occurs between children, though the vast majority of variance between districts can be partitioned into variances in agricultural input use and commercialization, healthcare access, and ethnic marginalization, combined with their estimated coefficients. The results generated by models designed to evaluate the robustness of the core estimation strategy also suggest that while variation among children is the largest source of variance, a small but nontrivial proportion of that variance is variance among households.
Degree
M.S.
Advisors
Shively, Purdue University.
Subject Area
Agricultural economics
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.