Inference Methods and Issues after Termination of a Sequential Trial
by
W. J. Hall (''Jack'')
University of Rochester

After completion of a standard sequential clinical trial, resulting in acceptance or rejection of a null hypothesis, a variety of further inference questions arise: estimation of the primary effect, testing secondary hypotheses, estimation of secondary parameters, evaluation of an expanded statistical model, evaluating interactions between treatment and strata or covariates, incorporating lagged data, and so on. We will review the considerable progress that has been made on some of these, and discuss some outstanding issues. Along the way, difficulties in applying concepts designed for non-sequential inference to a sequential setting will be identified and illustrated.

Date received: April 17, 2007