Decide first whether data are quantitative or qualitative. Unless mentioned otherwise,no assumption has been seriously violated, and, therefore, the appropriate test should be selected from among t, F, or 2 tests. Specify the precise form of the test. For example, specify that the t test is for two population means with related samples, or that the 2 test is for two variables, or that the F test is for repeated measures.

21.4 Another investigator criticizes the design of the previous study, saying that high school students should have been randomly assigned to either the special training condition or a control condition and tested just once at the end of the study. Subsequently, she conducts this study.