Overview of Statistical Tests Matching: For each of the following study situations, indicate the most appropriate statistical test by writing the letter in front of the test name in the space provided. If transformations are involved, your answer should include two letters (one for the transformation and one for the test). You may write a brief explanation of your choice in the space provided below each question to obtain partial credit, but this is not necessary. All options will NOT be used, and some could be used more than once. A. One-sample t-test Mann-Whitney U-test M. Pooled Variance 2-sample t-test G. B. Paired t-test H. Bonferroni procedure . Separate Variance 2-sample t-test C. Sign test 1. Analysis of Variance O. Power/Exponential transformation D. F-test J. Log/Root transformation P. x2 Test of Homogeneity E. Regression K. Pearson's Correlation Q. x2 Goodness-of-Fit Test F. Kruskal-Wallis test L. Spearman's Correlation R. X2 Test of Independence 1. To determine if resistance training can prevent or reverse loss of bone density in post-menopausal women, a research studied two groups of n = 50 women each. The treatment group participated in a standardized resistance training regime 3 days/wk for a year, while the control group did not. At the end of the study, bone density was measured. A boxplot of the data from the treatment group indicates the data distribution is negatively skewed, and includes one outlier that is smaller than the other values. The data from the control group is also moderately negatively skewed. The standard deviations are ST = 24 and Sc = 18. 2 . Suppose the investigator in the study described in (1) wanted to determine if the group of women who participated in the resistance training had greater variance in bone density than the women in the control group. Let's assume that the data distributions for both groups were Normally distributed. What test would be appropriate ? nulog 3. To determine if nutrient inputs to a lake from fertilizer runoff will increase algae growth, water from the lake is used to set-up 130 algae cultures. 65 cultures are randomly assigned to each of two groups, one receives a solution of commercial fertilizer dissolved in lake water, the other receives a similar volume of lake water with no fertilizer. After 1 week, algae population density is measured (cells/ml) in each culture. A boxplot for the fertilized group indicates data values are moderately positively skewed, and include one large outlier. The data for the control group are approximately Normally distributed. Standard deviations are: SF = 24 and Sc = 18. 4. Suppose you wanted to perform a study to determine which component of commercial lawn fertilizer might cause algae blooms in the lake. You use lake water to set-up 4 groups of algae cultures, each with n = 10 cultures. In Group 1 cultures you add no fertilizer, in Group 2 you add nitrogen, in group 3 you add phosphorous, and in group 4 you add potassium. After 1 week you get the results for mean algae biomass and standard deviations below, and exploratory data analysis indicates the data for the Phosphorous groups was moderately positively skewed. The data for the other groups were approximately Normally distributed, except the Control group had one outlier value smaller than the other data values. Group means and variances are listed below. Control Nitrogen Phosphorous Potassium X 225 353 600 237 S 501 636 enera 3225 404 oildug jenieneg 5. "Suppose that during experiment described in (4) two cultures for the phosphorous group and one culture for the control group were lost, but otherwise the results were the same. Which test would be appropriate
