1- A systematic error in the design or conduct of a study that results in an incorrect estimate of an exposure's effect on the risk of the outcome is called: a. Confounding b. Bias c. Stratification d. Effect modification 2- If a correlation were observed between per capita soft drink consumption and incidence of pancreatic cancer by country, this correlation, while possibly suggestive of a real association (that could be studied further with more analytic studies) would be subject to: a. The ecologic fallacy b. Differential recall bias c. Non-differential recall bias d. Effect modification 3- Either factor A or B can cause a certain disease, but only when followed by exposure to factor Z. Exposure to factor Z alone is not followed by disease, but the disease never occurs in the absence of exposure to factor Z. Factor Z is: a. Necessary and sufficient c. Sufficient but not necessary b. Necessary but not sufficient d. Neither necessary nor sufficient 4-Referring to the scenario above, factor B is: a. Necessary and sufficient c. Sufficient but not necessary b. d. Necessary but not sufficient Neither necessary nor sufficient 5- A hypothetical cohort study included 10,000 exposed and 10,000 unexposed people. Among the exposed, 500 got the disease. Among the unexposed, 100 got the disease. What is the relative risk of disease among those exposed relative to those not exposed? a. 4.0 b. 4.21 c. 5.0 d. 5.21 6- Referring to the scenario above assume that 20% of those who were exposed AND diseased were lost to follow-up. What would be the observed relative risk of disease among those exposed relative to those not exposed? a. 3.17 b. 4.00 c. 4.04 d. 4.17 7- A strategy which is not aimed at reducing selection bias is: a. Development of an explicit case definition c. Standardized protocol for structured interviews b. Encouragement of high participation rates c. Standardized protocol for structured interviews 8- The purpose of a double-blind study is to: a. Achieve comparability of cases and controls b. Reduce the effects of sampling variation c. Avoid observer and subject bias d. Avoid observer bias and sampling variation 9-Suppose that a case-control study is done in which the cases are more likely than the controls to remember and accurately provide information about their exposure. This is an example of: a. Non-differential misclassification bias c. Selection bias b. Differential misclassification bias d. Exclusion bias 10- The null hypothesis refers to the assumption that there is no difference between treatment/exposure groups with regard to the outcome of interest. If the true relative risk is 5 and the estimated relative risk is 2, the direction of bias would be: a. Toward the null b. Away from the null c. Switch over d. Cannot be determined 11- If the true relative risk is 0.7 and the estimated relative risk is 0.3, the direction of bias would be: a. Toward the null b. Away from the null c. Switch over d. Cannot be determined 12- Methods to reduce information bias include all the following except: a. Training of interviewers c. Blinding of interviewers b. Standard data collection form c. Consistent case definition 13- In order for a factor to be a confounder it must: a. be associated with the exposure c. not in the causal chain b. be associated with the outcome d. all of the above 14- Methods to control for a confounding variable in the design stage include all the following except: a. matching b. randomization c. restriction d. adjustment 15- Which of the following techniques can be used to control for confounding in the analysis stage of a study? a. randomization b. stratification c. effect modification d. blinding A hypothetical case-control study was done to examine the association between coffee consumption and anorexia among teenagers. The odds ratios, comparing the odds of frequent coffee consumption among teenagers who have anorexia to the odds of frequent coffee consumption among teenagers who do not have anorexia, are shown in the following table: Odds ratios Both sexes 5.1 Males 1.5 Females 1.5 16- Based only on these results, which of the following statements would apply: a. gender is a confounder c. the risk of anorexia is the same in males and females b. coffee consumption is a confounder d. gender is an effect modifier 17- Surgeons at Hospital A report that the mortality rate at the end of a 2-year follow-up after a new surgical procedure is 20%. At Hospital B, the surgeons report a 2-year mortality rate of 8% for the same procedure. Before concluding that the surgeons at Hospital B have vastly superior skill, which of the following possible confounders would you examine? a. b. c. d. The severity (stage) of disease of the patients at the two hospitals at baseline The starting point of the follow-up (after operation versus after discharge) The difference in the post-operative care at the two hospitals All of the above The commentary by Sir Austin Bradford Hill (1965), in which he describes criteria for judging whether an observed association is causal remains one of the most influential texts in the epidemiological literature. The guidelines offered by Gordis in the course text (pp. 236239) essentially restate those concepts. Bradford Hill's original criteria (and the corresponding contemporary guidelines of Gordis) are as follows: a. Strength (strength of the association) b. Consistency (replication of the findings) c. Specificity (specificity of the association) d. Temporality (temporal relationship) e. Biological gradient (dose-response relationship) f. Plausibility (biologic plausibility) g. Coherence (consistency with other knowledge) h. Experiment (cessation of exposure) i. Analogy (consideration of alternate explanations) For each definition identified in questions 18 through 21, indicate the Bradford Hill criterion it most closely describes. 18- The association can be explained in terms of known biologic mechanisms. a. Strength b. Consistency f. Plausibility c. Specificity g. Coherence d. Temporality h. Experiment e. Biological gradient i. Analogy 19- -Findings in a study indicated that the greater the exposure to tobacco the greater the risk of lung cancer. a. Strength b. Consistency f. Plausibility c. Specificity g. Coherence d. Temporality h. Experiment e. Dose Response i. Analogy 20- Exposure must precede the outcome. a. Strength b. Consistency f. Plausibility c. Specificity g. Coherence d. Temporality h. Experiment e. Biological gradient i. Analogy 21- An increase in the level, intensity, or duration of exposure is associated with increases in the magnitude of the risk. a. Strength b. Consistency f. Plausibility c. Specificity g. Coherence d. Temporality h. Experiment e. Biological gradient i. Analogy