Dynamic association modeling in 2×22×2 contingency tables

Association and dynamic association analyses in contingency tables, that take time into account, are of great interest in many disciplines. In this work, we define three dynamic association models: polynomial trend, Markovian, and sin–cos dependence models. We then discuss at length the classical and Bayesian analysis of these models. These results are then utilized to analyze postiche data, wherein we carry out the estimation of model parameters by using a genetic algorithm.