Consider the following scenario:

Psychological and social factors can influence the survival of patients with serious disease. One study looked at 54 patients and categorized them according to whether they owned a pet (Pet or No Pet). One year later the same patients were again categorized by whether they were still alive (Yes, No).

Suggested approach:

To see if there was any association between these variables it was proposed that first the data should be summarized using a contingency table to show the counts of patients in each category and then bar graphs produced to visualize the data distribution. Then, the means and standard deviations for each group should be calculated and regression analysis could then be used to explore the association between pet ownership and survival after 1 year.

Discuss the suggested approach to address this scenario. In particular,describe which parts of the approach are appropriate and/or inappropriate;and for any inappropriate analyses discuss why and possible alternatives.

• Do not perform any analysis for this question.
• Your response should be typed into the space below.