Discussion

of sample sizes between the two stages then becomes an interesting question. Of course, one must be very cautious about using optimality results that are based on unveriŽ able model assumptions, especially with binary data. Within the reliability literature, the generally accepted approach has been to strive for compromise test plans that achieve a reasonable trade-off between efŽ ciency and robustness/sensitivity to assumptions (see Meeker and Escobar 1998 and references therein). Investigation of similar plans for the present situation is a worthwhile research project.