GENERALIZED ORDERED LOGIT MODEL WITH TESTING ASSUMPTIONS: A CASE STUDY OF USING URBAN LIGHT RAIL IN BURSA

Abstract

Logistic regression has been a widely used prediction technique to analyze categorical variables. However, if the assumptions are violated the results may be biased. The study in this paper applied an analytical technique namely generalized ordered logit model. A case study of using urban light rail under pandemic conditions was applied for the analysis. The results suggested that logistic regression should not be applied before exploring the multicollinearity and applying the test of parallel lines. If the assumptions are violated, generalized ordered logit model should be considered. Regarding the predictive variables, sociodemographic, socioeconomic and travel pattern, related variables were found to have a statistically significant impact on the perception of safety and infrastructure of urban light rail. The outcomes of the study would provide a deeper understanding of developing regression models for categorical variables for future studies.