1 ANOVA is an appropriate statistical measure when we want to a) determine whether the hypothesized mean in one population is the same as another population mean. b) estimate the size of the difference between two or more group means. c) compare the means of three or more populations at once. d) compare the distributions of two or more categorical variables. Question 2 In ANOVA, we calculate an F statistic. The F statistic is the ratio of a) the difference in group means compared to the expected group difference in means. b) the variation within groups to the total variation. c) the variation in one group compared to the variation in the other. d) the variation between groups to the variation within groups. Question 3 What conditions must be met for the results of an ANOVA test to be reliable? Check all that apply. a) The variances for each population must be approximately equal to each other. b) The populations from which each group is drawn must be approximately Normal. c) The observations must be independent of each other and randomly selected from their populations. d) The hypothesized means of the populations from which each group is drawn must be approximately equal. Question 4 If the F statistic is significant, what is the next step in the analysis? a) Try to fit the best-fitting regression line. b) Re-run the F test to confirm the results. c) Run post-hoc tests to compare the group means. d) Calculate Cohen's d as a measure of effect size. Question 5 If the null hypothesis for an ANOVA test is A = B = C, what is the appropriate alternative hypothesis? a) At least one of the means is different. b) A B or A C c) A B C Question 6 Below is a source table for an ANOVA analysis. Assume the test is run with = 0.05. Fill in the missing values, using only the data provided in the table. Source SS df MS F statistic p-value Between 100 2 _____ __________ 0.002 Within 301 ____ _____ __________ _____ Total _____ 45 _____ __________ _____ (note: report your F statistic to 3 decimal places) Question 7 Based on these results, would you reject or fail to reject the null hypothesis? a) fail to reject b) cannot be determined c) reject Question 8 How many groups are being compared in this ANOVA? (Hint: look at df) a) 45 groups b) 3 groups c) 2 groups d) 47 groups Question 9 How many post-hoc pairwise comparisons would be required to compare the means of these groups? Remember: k(k1)2 a) 4 b) 2 c) 6 d) 3 Question 10 What significance level would you set for each post-hoc test to maintain an overall significance level of 0.05 if you are applying the Bonferroni correction? a) 0.05 b) 0.003 c) 0.15 d) 0.017