MULTIPLE HYPOTHESIS TESTS IN MULTIPLE INVESTIGATIONS

Inferential statistical methods have traditionally been based on the a ssumption that one experiment is performed and that interest centres o n one or more predetermined hypothesis tests. Exploratory research, on the other hand, often involves multiple hypotheses or repeated invest igations under similar or different conditions or both. Several techni ques have been proposed to deal with multiple or simultaneous hypothes is testing in single investigations, and procedures to combine observe d significance levels for an individual hypothesis test from two or mo re investigations have been suggested. In this paper we propose a meth od for identifying important results from multiple statistical tests i n multiple investigations. The method is illustrated by using high per formance liquid chromatography to identify potential aetiologic contam inants in L-tryptophan samples.