You hypothesize that practice tests will improve grades by 5 percent, but have no previous research to support your hypothesis. The mean score for students in general on test 1 is 64.0 (st dev = 12.1).

a) If you allow 49 students access to the practice tests what would be your power based on this information above if your hypothesis is correct?

b) Is this an experiment you want to carry out? Why or Why not? (You MUST provide statistical evidence to back up your decision, i.e., interpret your answer from a), above)

• You are studying the effect of bright-light therapy on depression. You have been recording the depression scores of all your clients for the past 25 years and the mean score was 66.4 (st dev = 13.1, higher numbers indicate more severe depression). You hypothesize that using bright-light therapy should reduce this severity of depression in your patients because research done in the past has shown consistently that use of birth lights reduces levels of depression. You run a small study with just 14 people who recently used bright-light therapy to estimate your effect size. You find that these 14 people had a mean score of 61.2 (st dev = 15.4). If this represents an accurate estimate, how many people would you have to include in your research to obtain a power of .80?