56
General Stopping Rule
•Analyses when sample sizes N1,…, NJ
–Can be randomly determined
•At jth analysis choose stopping boundaries
–aj < bj < cj < dj 
•Compute test statistic T(X1 ,…, XNj)
–Stop if T  < aj                     (extremely low)
–Stop if bj < T < cj       (approximate equivalence)
–Stop if T > dj                     (extremely high)
–Otherwise continue (with possible adaptive modification of analysis schedule, sample size, etc.)