Generalized Linear Model Menggunakan Distrbusi Lognormal dan Gamma: Aplikasi Terhadap Indeks Demokrasi Indonesia di Jawa Barat

Abstract

Generalized Linear Models (GLM) is a statistical approach that generalize linear regression models to handle response variables that are not necessarily normal. This research uses Gamma and Lognormal distributions to analyze the effect of Labor Force Participation Rate, Gini Ratio, and Percentage of Working Population on the Indonesia Democracy Index in West Java during 2021-2023. Both distributions are used and compared using AIC. Some of the results of this research are: 1) Lognormal GLM has a lower AIC, so it is better to use than Gamma GLM. 2) From these two models, obtained the same analysis results, that are Gini Ratio and Percentage of Working Population have a significant effect on Indonesia Democracy Index, with the Gini Ratio having a negative effect and the Percentage of Working Population having a positive effect. This shows that the higher of the Gini Ratio, the lower the Indonesia Democracy Index and vice versa, while the higher of the Percentage of Working Population, the higher the Indonesia Democracy Index and vice versa. 3) Meanwhile the Labor Force Participation Rate does not have a significant effect on the Indonesia Democracy Index. This research can be concluded that reducing social inequality and increasing population participation in employment contribute to improving the quality of democracy in West Java.