Determinants and Predictive Power of Fitted Model for Under-Five Child Mortality in Bangladesh

South Asian Journal of Population and Health, 10(1 & 2) 2017: 109-121

13 Pages Posted: 21 Feb 2018 Last revised: 16 Apr 2018

See all articles by A H M Musfiqur Rahman Nabeen

A H M Musfiqur Rahman Nabeen

University of Dhaka - Department of Statistics

Md. Abdus Salam Akanda

University of Dhaka- Department of Statistics

Date Written: December 31, 2017

Abstract

Under-5 mortality (U5M) is one of the central public health issues in many developing countries like Bangladesh. The main purpose of this study is to recognize determinants of U5M and to assess the predictive power of the fitted model. Univariate and multivariate logistic regression model is applied to identify the determinants using Bangladesh Demographic and Health Survey (BDHS), 2014 data. To assess the fitness of the model, Hosmer-Lemeshow Goodness-of-Fit test and Variance Inflation Factor (VIF) are used whereas Receiver Operating Curve (ROC curve) is used to measure the predictive power of the fitted model. The estimation results reveal that antenatal care, maternal education, exposure to media, mother’s age at birth, birth status, birth order number, caesarean delivery and wealth status are significant determinants of U5M. The model fitting tests suggest that the fitted model has an excellent predictive power. The identified potential determinants can be used in formulating policies to improve child survival in Bangladesh.

Keywords: Under-five mortality, Child survival, Predictive power, Odds ratio, Bangladesh

Suggested Citation

Nabeen, A H M Musfiqur Rahman and Akanda, Md. Abdus Salam, Determinants and Predictive Power of Fitted Model for Under-Five Child Mortality in Bangladesh (December 31, 2017). South Asian Journal of Population and Health, 10(1 & 2) 2017: 109-121. Available at SSRN: https://ssrn.com/abstract=3121951

A H M Musfiqur Rahman Nabeen

University of Dhaka - Department of Statistics

Dhaka, 1000
Bangladesh

Md. Abdus Salam Akanda (Contact Author)

University of Dhaka- Department of Statistics ( email )

Dhaka, 1000
Bangladesh

HOME PAGE: http://statdu.ac.bd/akanda/

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
13
Abstract Views
158
PlumX Metrics