Analysis of the generalized use of logit models in simulating changes in decision attributes at the airport check-in

Abstract

Logit is still very common in modal choice today, generally comparing different ways of transport considering attributes such as time and cost. This article compares and analyses the choices of check-in at airports. It considers factors of queue time and time of care, as presented in Negri et al. (2019). We made a Logit model with random coefficients which consider the heterogeneity of preference, when estimating the standard deviation according to the distribution assumed together with the coefficient. Therefore, the objective was to verify if the estimate with random parameters would provide extra information to analyst in the decision of check-in at airports. The study was performed with stated preference data for check-in in three main airports in Brazil. As a result, they indicated that by associating the surveys conducted in different airports, modeling the individuals’ preferences with traditional logit models allows losing generality. The logit model with random coefficients can capture differences in each individual. As example, in the analyze the queue time is understood more broadly than service time. The results indicate that the model with random coefficients brings information gain, as they manage to capture the differences among different people.