Expert Answers

Question 1
Comparison of mortality rates due to cancer of the uterus in users and nonusers of supplemental estrogen revealed the following results, where mortality rates are per 100,000: 3.0 (age 45-54) and 17.0 (age 55-70) among users of estrogen; 1.0 (age 45-54) and 6.0 (age 55-70) among nonusers of estrogen. A valid conclusion derived from the above data concerning mortality among estrogen users is:
The mortality rates for cancer of the uterus are lower in estrogen users than nonusers in both age groups studied. A causal relationship is demonstrated between the use of estrogen and incidence of uterine cancer.
Mortality from cancer of the uterus rises with age regardless of whether or not estrogen is used.
The mortality rate is lower in nonusers than users because the symptoms of uterine cancer are detected earlier in the former group of women.

Question 2
Controls are needed in a case-control study to evaluate whether the frequency of a factor or past exposure among the cases is different from that among comparable persons who do not have the disease under investigation.

Question 3
Selection bias is most likely to occur in which type of study?
Prospective cohort studies
Retrospective cohort studies
Case-control studies
Both retrospective cohort and case-control studies

Question 4
For a variable to be a confounder, it must be:
Associated with the disease (outcome) under study
Associated with the risk factor (exposure) under study
Both of the above
Neither of the above
Question 5
A study design method to control confounding involving the selection of study subjects so that potential confounders are distributed identically is called:
Multivariate modeling

Question 6
If a study fails to detect a significant difference between treatment A and treatment B when in fact the treatments do differ, the following type of error has occurred.
Type I error
Type II error
Treatment discrimination error
Difference testing error

Question 7
Major potential biases in a cohort study can result from each of the following except:
Loss to follow-up
Assessment of outcome

Question 8
When chance, bias, and confounding have been ruled out as alternative explanations for an association, we may conclude that the association is causal.
False .

Question 9
A retrospective cohort study compared to a prospective cohort study of the same kind of exposure can be completed more quickly.
False .

Question 10
Subjects with a confirmed diagnosis of carpal tunnel syndrome were compared to a comparable group without carpal tunnel syndrome, and both groups were asked about their prior occupational and recreational exposures, including hours per week of computer use. What type of study is this?
Case-control study
Prospective cohort study
Retrospective cohort study
Randomized clinical trial.

Question 11
Which of the following will not help to increase study precision (reduce random error)?
Using an accurate measuring instrument
Using an efficient study design
Using a small sample
Using repeat measures

Question 12
What sources can be used to identify controls for a case-control study?
General population
Hospital/clinic patients
Friends and relatives
All of the above

Question 13
A cross-sectional study allows the demonstration of a time sequence between the exposure and the outcome.
False .

Question 14
Loss-to-follow-up can be a problem in observational studies but not experimental ones.
False .

Question 15
When assessing a positive relationship between alcohol consumption and oral cancer using a case-control study, increasing the sample size of the study will result in which of the following?
A lower P value
A greater odds ratio
A smaller 95% confidence interval

A higher disease prevalence
The 1st and 3rd answers only
The 2nd and 4th answers only
The 1st, 2nd, and 3rd answers only
All of the above
None of the above

Question 16
Controlled clinical trials enable researchers to:
Derive knowledge about the origins of a disease
Control the level of exposure to a treatment
More accurately identify cause and effect
All of the above

Question 17
Randomized clinical trials can be used for each of the following except:
Evaluation of new drugs
Evaluation of alternate treatments
Assessment of screening programs
Identification of risk factors for disease

Question 18
The healthy worker effect is a form of what type of bias?
Selection bias
Control-selection bias
Interviewer/recording bias
Surveillance, diagnostic, or referral bias