Examples of Repeated Significance Tests
We consider below a number of examples of repeated significance tests in order to discover the extent to which the methods of Chapter IV can be adapted to a variety of more difficult problems. Often precise development of an asymptotic theory analogous to that obtained in Chapter IV for normally distributed data of known variability is complicated, and hence we shall concentrate on the use of normal approximations to obtain a rough idea of the probabilities and expected sample sizes involved. For continuously monitored tests Brownian motion provides simple approximations. For group tests one can use a discrete time normal approximation to take into account the sometimes substantial excess over the stopping boundary.
KeywordsBrownian Motion Confidence Region Nuisance Parameter Monte Carlo Experiment Repeated Significance
Unable to display preview. Download preview PDF.