ABSTRACT
In the field of cross-over designs, there has been a series of papers on the use of Bayesian inference by A. P. Grieve and co-workers. There is no doubt that Bayesian methodology will play an increasingly important role in the analysis of cross-over trials in the future, especially as software for carrying out the sometimes complex analyses becomes more readily available. This chapter illustrates one approach to dealing with repeated measurements within a cross-over trial. Finding the universally optimal cross-over design for any given class is complicated by the fact that the dependence between the responses in different periods on the same individual may be of difference forms, and the results are, in general, dependent upon a particular structure. The chapter also presents some closing thoughts on the key concepts discussed in the preceding chapters of this book.
