Abstract: This chapter discusses the multinomial response models. Multinomial response models, or multinomial logit models, are generalizations of logit models in which one or more independent variables are used to predict one or more polytomous dependent variables. If the dependent variable is dichotomous, then the multinomial response model is a logit model. Multinomial response models for two-way tables may be used to illustrate many of the general principles of multinomial response models. In the case of two-way tables, interpretation in terms of simultaneous logit models is attractive. A multinomial response model may be defined in terms of a series of related logit models. Multinomial response models for three or more dichotomous or polytomous variables are direct generalizations of multinomial response models for one polytomous dependent variable and one polytomous independent variable. In hierarchical models, a set of λ-parameters with a common superscript is either set to 0 or is completely unrestricted. Standardized parameter estimates are used to simplify models. Results of model simplification are judged in terms of the reduction achieved in the likelihood-ratio chi-square statistic.
Publication Year: 1979
Publication Date: 1979-01-01
Language: en
Type: book-chapter
Indexed In: ['crossref']
Access and Citation
Cited By Count: 1
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot