Introduction

Multinomial Discrete Choice Analysis

The MDC procedure provides maximum likelihood (ML) or simulated maximum likelihood estimates of multinomial discrete choice models in which the choice set consists of unordered multiple alternatives. The decision makers can be people, households, firms, or any other decision-making units, and the alternatives are a set of competing options. Unordered multiple choices are observed in many settings, including choices of housing location, occupation, political party affiliation, and mode of transportation.

The MDC procedure supports the following models and features:

  • intuitive

  • conditional logit

  • nested logit

  • heteroscedastic extreme value

  • multinomial probit

  • mixed logit

  • pseudorandom or quasi-random numbers for simulated maximum likelihood estimation

  • bounds imposed on the parameter estimates

  • linear restrictions imposed on the parameter estimates

  • SAS data set containing predicted probabilities and linear predictor (bold x prime bold-italic beta) values

  • decision tree and nested logit

  • model fit and goodness-of-fit measures, including the following:

    • likelihood ratio

    • Aldrich-Nelson

    • Cragg-Uhler 1

    • Cragg-Uhler 2

    • Estrella

    • adjusted Estrella

    • McFadden’s LRI

    • Veall-Zimmermann

    • Akaike’s information criterion (AIC)

    • Schwarz criterion or Bayesian information criterion (BIC)

Last updated: June 19, 2025