Pemilihan Model Regresi Logistik Ordinal Terbaik Menggunakan Metode Stepwise: (Studi Kasus: Data Indeks Prestasi Kumulatif Lulusan Program Sarjana FMIPA Unmul)

Selsi, Selsi
Darnah, Darnah
Wahyuningsih, Sri

ESTIMASI: Journal of Statistics and Its Application
Vol. 7, No. 1, Januari, 2026 : Estimasi

Ordinal logistic regression is one of the statistical methods used to model response variables with two or more categories that have levels. This study aims to model the cumulative grade point average data of undergraduate graduates of the Faculty of Mathematics and Natural Sciences, Mulawarman University in 2023 using ordinal logistic regression. Estimation of ordinal logistic regression model parameters is done using the Maximum Likelihood Estimation (MLE) method and Newton-Raphson iteration. The best model selection was conducted using the stepwise method based on the smallest Akaike Information Criterion (AIC) value and significant predictor variables. The selection process started with six predictor variables, then gradually eliminated three predictor variables because they were not significant and did not reduce the AIC value. The stepwise stage stopped at the model with three significant predictor variables that had an AIC value of 349,22. The results showed that the factors that had a significant effect on the cumulative grade point average of undergraduate graduates of the Faculty of Mathematics and Natural Sciences, Mulawarman University based on the best ordinal logistic regression model were study program, age, and admission pathway.