Mixed-effects multinomial logit model for nominal outcomes
Abstract
In this chapter, I first provide an overview of the classical multinomial logit regression model, and then specify the mixed-effects multinomial logit model. Next, a retransformation method is introduced that derives unbiased nonlinear predictions of the response probabilities by using analytic results of the mixed-effects multinomial logit model. The method is designed to retransform two random components, the between-subjects random effects and the within-subject random disturbances. The delta method is applied to compute the standard errors of the predicted probabilities based on the approximate variance-covariance matrix for a mean multinomial logit function. As ...
Get Methods and Applications of Longitudinal Data Analysis now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
