Power Activity: Dependent (Repeated) Measures The following two tables are only for DEPENDENT (Repeated) Measures. You'll see the word "Dependent" in the title of the table. Therefore, if you need Independent Measures tables, check the other set of tables in this worksheet. Also check each table to verify if it is estimating Power/EffectSize/SampleSize. Answer all "Dependent" questions using these two tables. You will have two questions per table. The scenarios for this worksheet will not be about dogs. Table 7-11 Approximate Power for Studies Using the / Test for Dependent Means for Testing Hypotheses at the .05 Significance Level Effect Size Difference Scores Small Medium Large in Sample (N) (d = .20) (d = .50) (d = .80) One-tailed test 10 15 46 78 20 .22 .71 .96 30 .29 .86 40 .35 .93 + + + 50 .40 .97 100 63 Two-tailed test 10 .09 .32 66 20 .14 .59 93 30 .19 .77 99 40 .24 .88 50 29 .94 100 55 *Power is nearly 1. Table 7-12 Approximate Number of Research Participants Needed for 80% Power for the Test for Dependent Means in Testing Hypotheses at the .05 Significance Level Effect Size Small Medium Large (d = .20) (d = .50) (d = .80) One-tailed 156 26 12 Two-tailed 196 33 14PART 1: Dependent (Repeated) Measures These first questions are about estimating how many participants you need in your study for obtaining a Power level of 80% (since that is the minimum desired for ALL items in this worksheet). (Table 7-12) 1. You want to do a study looking at the effects of a new anti-hunger weight loss pill. The participants will rate their hunger both before and after taking the pill. You will need to analyze your data using a dependent measures t-test, and have used a non-directional hypothesis (thus, you do a two-tailed test). You select your alpha level as a = .05. The effect size you are expecting, due to previous research, is d = .80. How many participants will you need to obtain a power level of 80%? 2. The study you completed in #1 was your pilot study. You want to do another study looking at the effects of the new anti-hunger weight loss pill. The participants will again rate their hunger both before and after taking the pill. You will still need to analyze your data using a dependent measures t-test, but have now used a directional hypothesis, stating that the participants will be less hungry when taking the pill (thus, you do a one-tailed test). You again select your alpha level as a = .05. The effect size you obtained in your pilot study was d =.50. How many participants will you need to obtain a power level of 80%? These next questions are about estimating how much power your study had based on the obtained effect size of your study and how many participants you had per group. (Table 7-11) 3. You do a study looking at the effects of a new anti-hunger weight loss pill. The 20 participants rated their hunger both before and after taking the pill. You analyzed your data using a dependent measures t-test, and used a non-directional hypothesis (thus, you did a two-tailed test). You selected your alpha level as a = .05. The effect size you obtained was moderate (medium) at d = .50. How much power did your study have? Is this an acceptable amount of power to confidently reject the null (and say there truly is a difference)? If yes, good. If no, how would you fix it (be specific and include your goal sample size)? 4. The study you completed in #3 was your pilot study. You replicate that study but were only able to obtain 30 participants. The participants again rated their hunger both before and after taking the pill. You analyzed your data using a dependent measures t-test, but due to the results last time, and justifiably expecting the same reduction in hunger, you used a directional hypothesis, stating that the participants will be less hungry when taking the pill (thus, you did a one-tailed test). You again selected your alpha level as a = .05. The effect size obtained in your replication study was again d= .50. How much power did your study have? Is this an acceptable amount of power to confidently reject the null (and say there truly is a difference)? If yes, good. If no, how could you fix it