Power and Effect Size 1. An article I found involving youths in treatment planning produced higher treatment success scores (M = 13.2) than the usual procedure of staff-only treatment planning (M = 8.2), or no treatment (M = 4.1), F(2,78) = 9.00, p = .02. a. What are the chances that there is really no treatment effect, and this statistical conclusion is a mistake? b. Describe any other type(s) of decision error(s) might these researchers have made? Specifically, what would that error be? c. Should I worry about the statistical power of this study? Why or why not? d. I'd like to compare the results of this study to another I read previously. Would you please compute the size (r) of this effect (r = [F/F+dferror])? Should I consider this a small, medium or large effect? e. I'm thinking of replicating this study, if I wanted to take only a 10% risk of getting a H0: when there really is an effect, how many youths should I have in each condition of the study? 2. A different study of the same youth-involved treatment procedure showed a dramatic difference between adolescent's scores on the Situation-Appropriate Behavior Scale at intake