ABSTRACT
This chapter considers an issue that arises in testing several contrasts. Namely, any time multiple tests are performed, the probability of committing a Type-I error increases for the collection of tests. The chapter presents methods that control for multiple testing and thus prevent inflation of the Type-I error rate. It approaches this issue primarily from the perspective of hypothesis testing. The chapter describes how to perform a variety of tests, and delineates situations where different types of techniques are appropriate. It presents a brief overview of the multiple-comparisons procedures. First, when a researcher plans to test a small number of contrasts based on theoretical hypotheses prior to data collection, a technique known as the Bonferroni adjustment is appropriate. Second, when all pairwise comparisons are of potential interest, Turkey developed a procedure to maintain αEW at.05. Third, Scheffe's procedure can be used when an investigator decides to test complex comparisons suggested by the data.
