Abstract
In the logistic regression problem and the generalized linear regression problem, the dispersion around the fitted regression is sometimes too large. Over-dispersion and random effect models are proposed to provide fitted regressions which conform more closely to the data. However, in the analysis of contingency tables the random effect cannot be detected or corrected, so it is important to understand the influence of the heterogeneity factor on contingency tables. Therefor, this paper investigates the influence of heterogeneity on binary data and evaluates its effect on contingency tables.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have