chapter  10
51 Pages

A Test Case Study

WithDavid A. Hensher, Lester W. Johnson, J.J. Louviere, J. Horowitz

This chapter aims to introduce a computer program to estimate the coefficients of the basic multinomial logit (MNL) model, and provide suitable output such as probabilities of selecting alternatives, direct and cross elasticities of choice, and inclusive value. The data set contains a detailed set of variables defining the levels of service associated with trips where the car and train are used as the main means of transport for the greater part of the work trip together with socioeconomic data. The chapter begins with general information on program control, and explains the three divisions of the program input commands - the general program control division, the data division, and the model division. In concluding the discussion on using BLOGIT, present the set up of the card deck that the analyst can prepare as input for the test run; and the output.