Many research projects involve testing multiple research hypotheses. These research hypotheses could be evaluated using comparisons of means, bivariate correlations, regressions, and so forth, and in fact most studies consist of a mixture of different types of test statistics. An important consideration when conducting multiple tests of significance is how to deal with the increased likelihood (relative to conducting a single test of significance) of falsely declaring one (or more) hypotheses statistically significant, titled the multiple comparisons problem. This multiple comparisons problem is especially relevant to the topic of research design because the issues associated with the multiple comparisons problem relate directly to designing studies (i.e., number and nature of variables to include) and deriving a data analysis strategy for the study. This entry introduces ...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles