Chapter 1
Exercises 6 (page 19)

Yes, this is a confounding factor. The pharmaceutical companies can skew their study group to be favorable to the result that they would like to see. The subjects that the company chooses to use in the study may be a confounding factor. If a randomized group of subjects are chosen and a treatment and control group is used in the study then some bias may be reduced that is due to confounding factors.

13 (page 25)

Occupation
Total Death
Total Employed
% death of total employed (total death/total employed)
Driver/sales workers and truck drivers
993
3,412,370
0.029%
farmers and ranchers
341
829,680
0.041%
construction laborers
339
1,493,390
0.023%
miscellaneous agri workers
176
758,620
0.023%
aircraft pilots
81
121,070
0.067%
logging workers
80
86,110
0.093%
fishers and fishing workers
48
40,540
0.118%
electrical power line installers/repairers
36
110,090
0.033%
structural iron and steel workers
35
62,940
0.056%
refuse and recyclable material collectors
32
73,050
0.044%

It seem as though fishers and fishing workers is the occupation with the most deaths per total workers versus construction laborers and miscellaneous agricultural workers who seem to have the least dangerous jobs.
However, the data does not state the time each occupation spends on their jobs. Farmers, ranchers, fishers and truck drivers may be spending many more hours working in dangerous conditions as company to electrical power line installers. Therefore, the work condition of each occupation should be taking into consideration. Furthermore, the age and health conditions of the employees should also be considered. If one occupation tend to have elder workers then it is likely that some of these employees may die because of their inherent age and not their occupation.

15 (page 26)

a) This study is observational.
b) Some subject may just inherently want to make the decision and be done with the study, whereby others decided to deliberate and answer later. There is no control on when the candidates make the