Exercises 13d. How might Beth go about “random sampling”? How would she go about “random as-
signment”?
e. If random assignment is not possible in this study, does that have negative implications
for the validity of the study?
f. What are some of the variables that might influence the outcome of this study separate
from any true population differences between boys’ and girls’ incomes?
g. Distinguish clearly between the descriptive and inferential statistical features of this ex-
ample.1.21 The Journal of Public Healthpublished data on the relationship between smoking and
health (see Landwehr & Watkins [1987]). They reported the cigarette consumption per adult
for 21 mostly Western and developed countries, along with the coronary heart disease rate
for each country. The data clearly show that coronary heart disease is highest in those coun-
tries with the highest cigarette consumption.
a. Why might the sampling in this study have been limited to Western and developed
countries?
b. How would you characterize the two variables in terms of what we have labeled “scales
of measurement”?
c. If our goal is to study the health effects of smoking, how do these data relate to that
overall question?
d. What other variables might need to be considered in such a study?
e. It has been reported that tobacco companies are making a massive advertising effort in
Asia. At present, only 7% of Chinese women smoke (compared with 61% of Chinese
men). How would a health psychologist go about studying the health effects of likely
changes in the incidence of smoking among Chinese women?