# Question

Periodontal disease may play a role in many diseases, some of which were unknown previously. For example, a recent study of the failure of joint replacement prostheses due to aseptic loosening demonstrated a link with bacterial DNA that was also found in dental plaque. Therefore, it is crucial that methods of determining bacterial DNA in plaque have a high degree of reliability. A study was conducted to examine the variability in the chemical analyses for specified bacterial DNA content in plaque. The two major sources of variation selected for investigation were the person con-ducting the analysis and the subjects supplying the plaque. The researchers randomly selected five analysts from a large pool of experienced analysts and 10 female subjects (ages 18– 20). Plaque was scraped from the entire dentition of each subject and divided into five samples. Each of the analysts was given an unmarked sample from each of the subjects. The analysts then made a determination of the DNA content (in micrograms) for each of the 10 samples. The data are given here.

a. Why do you think the researchers selected only female subjects who were essentially the same age?

b. Write an appropriate linear statistical model identifying all terms in the model.

c. State the null and alternative hypotheses for testing for an effect due to analyst.

a. Why do you think the researchers selected only female subjects who were essentially the same age?

b. Write an appropriate linear statistical model identifying all terms in the model.

c. State the null and alternative hypotheses for testing for an effect due to analyst.

## Answer to relevant Questions

Refer to Exercise 17.5. a. Write down the expected mean squares. b. Run an analysis of variance and test for a significant analyst effect. c. Estimate the variance components associated with analysts, subjects, and error. ...Refer to the data from Exercise 18.9. Give a model for this design, and run a repeated measures analysis of variance to compare the two treatment groups. Do the analysis of variance results agree with your intuition based on ...Refer to Exercise 18.13. In spite of the results from period 2, we can still get a between patient comparison of the treatment groups if we use the period 1 results only. Suggest an appropriate test, run the test, and give ...Refer to Exercise 18.21. Using just the period 1 data, test for a difference in the fourformulations’ mean AUCs. Are your results consistent with the conclusions from Exercise 18.22? Why might the analysis of Exercise ...Refer to Exercise 18.30. Consider as your response variable the proportional change in the mean percentage of leakage at time 3 hours and at time 0. That is, y = P3 – P0/P0 where P0 and P3 are the percentage of leakage ...Post your question

0