Department of Statistics, Shahid Beheshti University, Tehran, Iran
Abstract
Discrete mixed response analysis is an important statistical topic in various branches of science. Discrete variables include ordinal variables and overdispersed binomial variables. Overdispersed binomial data is the sum of uncorrelated Bernoulli trials with the probability of success. In this paper, a random-effects model is introduced for analyzing overdispersed and ordinal binomial mixed responses under longitudinal study. In this model, the overdispersed binomial response variable is assumed to follow the beta-binomial distribution and the latent variable approach is used to model the ordinal response variable. Also, the model parameters are estimated using the maximum likelihood method and the parameter estimates are evaluated using the Monte Carlo simulation method. Finally, the application of the introduced model to real data is examined.
Sedighe Azimi, S. & Bahrami Samani, E. (2020). Combined beta-binomial and ordinal model with random effects for longitudinal mixed response analysis. Andishe_ye Amari, 24(2), 9-14.
MLA
Sedighe Azimi, S., & Bahrami Samani, E. "Combined beta-binomial and ordinal model with random effects for longitudinal mixed response analysis", Andishe_ye Amari, 24, 2, 2020, 9-14.
HARVARD
Sedighe Azimi S., Bahrami Samani E. (2020). 'Combined beta-binomial and ordinal model with random effects for longitudinal mixed response analysis', Andishe_ye Amari, 24(2), pp. 9-14.
CHICAGO
S. Sedighe Azimi & E. Bahrami Samani, "Combined beta-binomial and ordinal model with random effects for longitudinal mixed response analysis," Andishe_ye Amari, 24 2 (2020): 9-14,
VANCOUVER
Sedighe Azimi S., Bahrami Samani E. Combined beta-binomial and ordinal model with random effects for longitudinal mixed response analysis. Andishe_ye Amari. 2020;24(2):9-14 (In Persian).
