Regression approach to analyze the travel characteristics of university students

Abstract

This study at the University of Engineering and Technology (UET), Lahore, examined students’ travel behavior, mode choice and constraints. Data from 1,449 students were analyzed using various statistical methods, including descriptive analysis, latent class analysis (LCA), multinomial logistic regression (MNL), correlation, and linear regression. The model’s accuracy was evaluated by calculating the percent error. Walking was the most popular mode of transportation within the campus, preferred by 37% of students. Most students traveled 1-15 km, taking 10-15 minutes, with off-campus students typically making three trips to their destination. Safety concerns significantly influenced 37% of students’ travel behavior due to built-in constraints.