Unformatted text preview: Assignment #5
(Due: March 25, 2011)
1. An experiment is conducted to compare treatments for cholesterol reduction. Forty patients (40) with high cholesterol are recruited and randomly assigned to treatment groups. Some of them are in the Toledo area and others are in Cincinnati. They are given unique ID numbers for identification. At the start of the study they take surveys to record some lifestyle variables, and the initial cholesterol measurements are taken. At the end of the treatment period, cholesterol measurements are taken and the reduction in cholesterol is recorded. There were some dropouts along the way so there is not complete data for everyone. Save only those with complete information. The survey data may be found in the file cholc.txt, and the cholesterol measurements are in chola.txt for Toledo and cholb.txt for Cincinnati. The researchers hire you to analyze the data. a. Convert each .txt file into a SAS data. Introduce a categorical variable called clinic while converting chola.txt and cholb.txt, which helps distinguish the data for Toledo and Cincinnati, respectively. Hints: (1) Note that Chola.txt has ID information for two people in the same row. For this, after a usual “infile” statement, use input ID TRT $ CHOLRED @@; (2) The data cholb.txt gives cholesterol level before and after. So, in your data step for converting cholb.txt, you need to introduce a new variable CHOLRED = before – after. Then, drop the variables before and after. b. Combine all these three SAS datasets into a single SAS dataset. Call the combined data chol. Delete missing observations for CHOLRED. Now answer the following questions for the researchers. For each of the questions, use the p‐value to test the hypotheses. Use level of significance α = 0.05 throughout to make a decision. c. What are the average cholesterol reductions for each treatment? Are these differences statistically significant? Which treatment has the highest average cholesterol reduction? Use appropriate tests. d. The researchers are also interested in the effect of exercise on cholesterol reduction. Conduct a statistical test to see if there are any significant differences in cholesterol reduction according to whether the patient exercises regularly. What about smoking? Do smokers have different cholesterol reduction, on average, than nonsmokers? Explain. e. Instead of answering the previous questions in part e separately, run a two‐way ANOVA model with interaction for the same questions and interpret the result. Also, draw an interaction plot using the mean CHOLRED for each combination of smoke and exercise. g. Are smokers just as likely to exercise regularly as nonsmokers? Explain. ...
View Full Document
This note was uploaded on 06/06/2011 for the course STAT 4360 taught by Professor Park during the Spring '11 term at UGA.
- Spring '11