Factorial design for a sequential allocation study

Abstract

Robert J. Taylor and Herbert A. David developed a new experimental approach in the screening of prospective drugs to be used in the treatment of cancer. The procedure is a sequential one, where the allocation of patients to drugs is based upon the performance of the drugs in immediately previous periods. Taylor further developed a computer program to study the effectiveness of the procedure by means of simulated trials.

The purpose of this paper has been to study the sequential allocation procedure further by means of simulation in the form of a factorial experiment. Of primary concern has been the behavior of different weighting functions operating under varying experimental conditions. Response variables have been studied as a means for evaluating the effectiveness of the procedure. The results of this experiment are comparable with the findings originally presented by Taylor and David.