Based on Schachter’s findings, we might suggest a new hypothesis. Perhaps people want to be with others when they are anxious because being with others causes them to become less anxious. How would we go about testing this hypothesis? We might place subjects in one of two conditions. In one condition, subjects are asked to spend 20 minutes waiting in a room with another person. At the end of the waiting period, the subjects’ anxiety levels are measured. If the hypothesis is correct, subjects who wait alone should be more anxious than subjects who wait with another person. The independent variable in this experiment is affiliation. We manipulate affiliation by assigning subjects to wait either alone or with others. The dependent variable is anxiety. According to the hypothesis, anxiety depends on whether subjects wait alone or with another person. In Schachter’s original experiment, anxiety was the independent variable, and affiliation was the dependent variable. As you can see, we changed the status of these variables when we modified the hypothesis.