ABSTRACT
This chapter presents the sequential statistical approaches to interaction in a way that is relatively free of complex terminology and intimidating notation. When sequential problems are cast in log-linear terms, their analysis becomes almost straight forward. Moreover, the basic approach should seem familiar to researchers previously schooled in traditional approaches to data analysis and hypothesis testing. We have distinguished between event and time sequences here because the terms appear frequently in the literature and because organizing the examples this way makes it easier, we think, for readers to recognize in them something that applies to their own work. Comparison of chi-square, log-linear, and logit approaches, stressing their common features. We demonstrate further how a number of common problems in sequential analysis can be solved by applying logit models to the appropriate contingency tables.
