New general location models for mixed responses

نویسندگان

1 Department of Statistics, Faculty of Mathematical Science, Shahid Beheshti University, Tehran, Iran

doi
10.22060/ajmc.2025.23674.1284
چکیده

In this paper, we introduce new general location model for mixed responses including correlated nominal, ordinal and continuous outcomes by using latent variable approach. We discuss regression methods for jointly analysis of continuous and categorical (nominal and ordinal) responses. After presenting the Leon and Carrière general location model [7], new general location model is introduced. A full likelihood-based approach is used to obtain maximum likelihood estimations of the models parameters. The proposed model is applied to BMI, Steatosis and Osteoporosis data.