ABSTRACT

Researchers often need to collect measurements on pairs of variables because they want to study their relationship. For example, in the section on pages 146-149 we described a study where subjects were given words and pictures to commit to memory. It appeared that the mean number of pictures remembered was significantly larger than the mean number of recalled words. Thus, different means corresponded to different levels of the independent variable (i.e., treatments). Looking at this result from a slightly different perspective, we can say that the means were related to the treatments. This example indicates that speaking about differences between treatment means is the same as saying means are related to treatments.