Think about the relationship between the independent variable, whether or not people have knowledge of an election's predicted outcome (they either "know" or "don't know" the predicted outcome) and the dependent variable (they either "voted" or they "did not vote"). The reformers' idea links one value of the independent variable, "know the predicted outcome," with one value of the dependent variable, "did not vote." For the reformers' idea to qualify as an acceptable explanation, what else must it describe?