INVERSE GAUSSIAN REGRESSION MODELING AND ITS APPLICATION IN NEONATAL MORTALITY CASES IN INDONESIA
Abstract
Inverse Gaussian Regression (IGR) is a suitable model for modeling positively skewed response data, which
follows the inverse Gaussian distribution. The IGR model was formed from the Generalized Linear Models (GLM).
This study aims to model the IGR with applied to model the factors influencing the infant mortality cases of provinces
in Indonesia. Estimation of the IGR model parameters was employed by the Maximum Likelihood Estimation (MLE)
and Fisher scoring methods. The Likelihood Ratio Test (LRT) and Wald test were used for hypothesis testing of
significance parameters. The IGR model was applied to the infant mortality cases of provinces in Indonesia in 2020.
The data for modeling infant mortality cases using IGR were obtained from the Ministry of Health of the Republic of
Indonesia and the Central Bureau of Statistics. The result shows that the factors influencing the infant mortality cases
of provinces in Indonesia based on the IGR model were: the percentage of pregnant women who received bloodboosting tablets, the percentage of low birth weight, the percentage of complete neonatal visits (KN3), the percentage
of toddlers who received early initiation of breastfeeding, the percentage of toddlers who are exclusively
breastfeeding, the percentage of toddlers who received complete primary immunization, the percentage of households
with access to adequate drinking water, and the percentage of households with access to appropriate sanitation.