Multi-level models can benefit from minimizing higher-order variations: an illustration using child malnutrition data

Sumon Das, Azizur Rahman, Ashraf Ahamed, Sabbir Tahmidur Rahman

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

This study aims to measure the robustness of multi-level models designed for three anthropometric indices – height-for-age (HAZ), weight-for-age (WAZ) and weight-for-height (WHZ) Z-scores for estimating the childhood malnutrition indicators stunting, underweight and wasting in Bangladesh. The 2011 BDHS child malnutrition data have been used in developing multi-level models with and without incorporating specific contextual variables relating to lower administrative units extracted from the 2011 Bangladesh Population and Housing Census. The robustness of the models is examined through (i) testing significance of random effects corresponding to lower administrative units through selection criteria including conditional AIC, R-squared, and LRT; (ii) comparing multi-level model-based estimators to design-based estimators of child malnutrition indicators with their precision at division, district and sub-district levels; and (iii) assessing the impact of contextual variables in capturing higher-order administrative level variations. Findings reveal that the inclusion of important contextual variables helps capture variations at higher-level administrative units, and consequently assists in the selection of robust multi-level models which ultimately provide improved accuracy of estimated parameters. The findings support the application of lower administrative census information in developing a simpler multi-level model by minimizing higher-order variation.
Original languageEnglish
Pages (from-to)1090-1110
Number of pages21
JournalJournal of Statistical Computation and Simulation
Volume89
Issue number6
DOIs
Publication statusPublished - 13 Apr 2019

Keywords

  • Contextual variable
  • conditional AIC
  • child anthropometric indices
  • population hierarchy
  • random effects model
  • POVERTY
  • DETERMINANTS
  • LEVEL
  • SELECTION

Cite this

@article{57e326feb3c84c1ab781a8322e825e9b,
title = "Multi-level models can benefit from minimizing higher-order variations: an illustration using child malnutrition data",
abstract = "This study aims to measure the robustness of multi-level models designed for three anthropometric indices – height-for-age (HAZ), weight-for-age (WAZ) and weight-for-height (WHZ) Z-scores for estimating the childhood malnutrition indicators stunting, underweight and wasting in Bangladesh. The 2011 BDHS child malnutrition data have been used in developing multi-level models with and without incorporating specific contextual variables relating to lower administrative units extracted from the 2011 Bangladesh Population and Housing Census. The robustness of the models is examined through (i) testing significance of random effects corresponding to lower administrative units through selection criteria including conditional AIC, R-squared, and LRT; (ii) comparing multi-level model-based estimators to design-based estimators of child malnutrition indicators with their precision at division, district and sub-district levels; and (iii) assessing the impact of contextual variables in capturing higher-order administrative level variations. Findings reveal that the inclusion of important contextual variables helps capture variations at higher-level administrative units, and consequently assists in the selection of robust multi-level models which ultimately provide improved accuracy of estimated parameters. The findings support the application of lower administrative census information in developing a simpler multi-level model by minimizing higher-order variation.",
keywords = "Contextual variable, conditional AIC, child anthropometric indices, population hierarchy, random effects model, POVERTY, DETERMINANTS, LEVEL, SELECTION",
author = "Sumon Das and Azizur Rahman and Ashraf Ahamed and Rahman, {Sabbir Tahmidur}",
note = "data source: Data extracted from Bangladesh Demographic and Health Survey 2011 have been used as the survey data and contextual variables have been extracted from census reports of Bangladesh Population and Housing Census 2011",
year = "2019",
month = "4",
day = "13",
doi = "10.1080/00949655.2018.1553242",
language = "English",
volume = "89",
pages = "1090--1110",
journal = "Journal of Statistical Computation and Simulation",
issn = "0094-9655",
publisher = "TAYLOR & FRANCIS LTD",
number = "6",

}

Multi-level models can benefit from minimizing higher-order variations: an illustration using child malnutrition data. / Das, Sumon; Rahman, Azizur ; Ahamed, Ashraf ; Rahman, Sabbir Tahmidur .

In: Journal of Statistical Computation and Simulation, Vol. 89, No. 6, 13.04.2019, p. 1090-1110.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Multi-level models can benefit from minimizing higher-order variations: an illustration using child malnutrition data

AU - Das, Sumon

AU - Rahman, Azizur

AU - Ahamed, Ashraf

AU - Rahman, Sabbir Tahmidur

N1 - data source: Data extracted from Bangladesh Demographic and Health Survey 2011 have been used as the survey data and contextual variables have been extracted from census reports of Bangladesh Population and Housing Census 2011

PY - 2019/4/13

Y1 - 2019/4/13

N2 - This study aims to measure the robustness of multi-level models designed for three anthropometric indices – height-for-age (HAZ), weight-for-age (WAZ) and weight-for-height (WHZ) Z-scores for estimating the childhood malnutrition indicators stunting, underweight and wasting in Bangladesh. The 2011 BDHS child malnutrition data have been used in developing multi-level models with and without incorporating specific contextual variables relating to lower administrative units extracted from the 2011 Bangladesh Population and Housing Census. The robustness of the models is examined through (i) testing significance of random effects corresponding to lower administrative units through selection criteria including conditional AIC, R-squared, and LRT; (ii) comparing multi-level model-based estimators to design-based estimators of child malnutrition indicators with their precision at division, district and sub-district levels; and (iii) assessing the impact of contextual variables in capturing higher-order administrative level variations. Findings reveal that the inclusion of important contextual variables helps capture variations at higher-level administrative units, and consequently assists in the selection of robust multi-level models which ultimately provide improved accuracy of estimated parameters. The findings support the application of lower administrative census information in developing a simpler multi-level model by minimizing higher-order variation.

AB - This study aims to measure the robustness of multi-level models designed for three anthropometric indices – height-for-age (HAZ), weight-for-age (WAZ) and weight-for-height (WHZ) Z-scores for estimating the childhood malnutrition indicators stunting, underweight and wasting in Bangladesh. The 2011 BDHS child malnutrition data have been used in developing multi-level models with and without incorporating specific contextual variables relating to lower administrative units extracted from the 2011 Bangladesh Population and Housing Census. The robustness of the models is examined through (i) testing significance of random effects corresponding to lower administrative units through selection criteria including conditional AIC, R-squared, and LRT; (ii) comparing multi-level model-based estimators to design-based estimators of child malnutrition indicators with their precision at division, district and sub-district levels; and (iii) assessing the impact of contextual variables in capturing higher-order administrative level variations. Findings reveal that the inclusion of important contextual variables helps capture variations at higher-level administrative units, and consequently assists in the selection of robust multi-level models which ultimately provide improved accuracy of estimated parameters. The findings support the application of lower administrative census information in developing a simpler multi-level model by minimizing higher-order variation.

KW - Contextual variable

KW - conditional AIC

KW - child anthropometric indices

KW - population hierarchy

KW - random effects model

KW - POVERTY

KW - DETERMINANTS

KW - LEVEL

KW - SELECTION

U2 - 10.1080/00949655.2018.1553242

DO - 10.1080/00949655.2018.1553242

M3 - Article

VL - 89

SP - 1090

EP - 1110

JO - Journal of Statistical Computation and Simulation

JF - Journal of Statistical Computation and Simulation

SN - 0094-9655

IS - 6

ER -