||Correlated data are commonly encountered in many fields. The correlation may come from the genetic heredity, familial aggregation, environmental heterogeneity, or repeated measures. Royall and Tsou (2003) proposed a parametric robust likelihood technique. With large samples, the adjusted binomial likelihood is asymptotically legitimate for correlated binary data.|
In this work, we use the adjustment by the binomial working model and obtain a new method for estimating the correlation between data in a cluster.
||. Ahn, H. and Chen, J.J. (1997). Tree structured logistic model for over dispersed binomial data with application to modeling developmental effects. Biometrics, 53, 435-455.|
. Alan Agresti. (1990). Categorical data analysis. New York : Wiley, 1st ed.
. Bahjat F. Qaqish. (2003). A family of multivariate binary distributions for simulating correlated binary variables with specified marginal means and correlations. Biometrika, 90, 2, 455-463.
. Chen, C.H. and Tsou, T.S. (2007). Parametric robust inference about regression parameters for the correlation coefficient. Statistics, 41, 1-9.
. L. Blizzard and D. W. Hosmer (2006). Parameter estimation and goodness-of-fit in log binomial regression. Biometrical Journal 48, 1, 5–22
. Royall, R.M. and Tsou, T.S. (2003). Interpreting statistical evidence using imperfect models: robust adjusted likelihood functions. Journal of the Royal Statistical Society, Series B 65, 391-404.
. Terra L. Slaton, Walter W. Piegorsch and Stephen D. Durham (2000). Estimation and testing with overdispersed proportions using the beta-logistic regression model of Heckman and Willis. Biometrics 56, 125-133.
. Tsou, T.S. (2005). Inference of variance functions- a parametric robust way. Journal of Applied Statistics, 32, 785-796.
. Tsou, T.S. (2006a). Regression of correlated binary data – a parametric robust approach that automatically accommodates overdispersion and underdispersion.
. Tsou, T.S. (2006b). Robust Poisson regression. Journal of Statistical Planning and Inference, 136, 3173-3186.
. Tsou, T.S. (2007). A simple and exploratory way to determine the mean-variance relationship in generalized linear models. Statistics in Medicine, 26, 1623-1631.