In the previous chapter, I introduced analysis of variance, a statistical procedure frequently used to test whether two or more group means are statistically different. In the field of communication and other social sciences, the groups being compared are often created by the researcher in an experimental context. One of the methodologically elegant features of an experiment is its ability to roughly equate a set of groups on all individual differences that might be related to the dependent variable through the random assignment of the study participants to the experimental conditions. Because the participants in an experiment are assigned to groups at random, we know that the groups of people assigned to different experimental conditions will be, on average, roughly equivalent on all variables except the manipulated independent variable-the variable which is under the control of the researcher-prior to the beginning of the study.