A psychologist is interested in determining whether a new treatment for depression improves (decreases) depression scores. She randomly assigns 30 subjects to the treatment and 30 subjects to the control. At the conclusion of the study she uses an independent t-test to evaluate whether the groups differ on a widely used depression scale that ranges between 0 and 30. The treatment group mean is 18 and the control group mean is 22. What is the null hypothesis for the analysis