A study was done to identify factors affecting physicians’ decisions to advise or not to...

A study was done to identify factors affecting physicians’ decisions to advise or not to advise patients to stop smoking (Cummings et al. 1987). The study was related to a training program to teach physicians ways to counsel patients to stop smoking and was carried out in a family practice outpatient center in Buffalo, New York. The study population consisted of the cigarette-smoking patients of residents in family medicine seen in the center between February and May 1984.

a. We first consider whether certain patient characteristics are related to being advised or not being advised. The following table shows a breakdown by sex:

What proportion of the males were advised to quit and what proportion of the females were advised? What are the standard errors of these proportions? What is the standard error of their difference? Test whether the difference in the proportions is statistically significant.

Next consider a breakdown by race: White and Other versus African- American:

What proportions of the patients of male and female physicians were advised? What are the standard errors of the proportions and their difference? Is the difference statistically significant? The following table shows the breakdown according to whether the physician smokes:

Of those patients who sawa smoking physician, what proportion were advised to quit, and of those who saw a nonsmoker, what proportion were so advised? What are the standard errors of the proportions and of their difference? Is the difference statistically significant? Finally, this table gives a breakdown by age of physician:

What are the proportions advised to quit in each of the three age categories and what are their standard errors? Are the differences statistically significant?