21- A single study can have a variable that is both a confounder and an effect modifier
True or false
22-wich of the following will not help to increase study precision (reduce random error)?
-using an accurate measuring instrument
-using an efficient study desing
-using a small sample
-using repeat measures
23. Wich of the following statemets is true about the confidence interval?
- If you did the study 100 times and got 100 confidence intervals, in 95 of the 100 results, the true point estimate would lie within the givin interval.
- Assuming no bias or counfunding, the range within which the true measure of effect lies a stated probability, or a certain degree of assurance (usually 95%).
- The confidence interval is caculated around the point estimate, and quantifies the variability around the point estimate.
- All of the above.
24. The following study was highlighted in The New York Times on 12 Dec 2008: “Fort he sake heart disease research, 809 members of the Old Order Amish community agreed to go to a clinic in Lancaster, PA, near their homes, and drink a rich milkshake that was made mostly of heavy cream. Over the next six hours, a group of investigators took samples of their blood, determining how much fat was churning through their bloodstreams. Most of the study participants responded as expected – their levels of triglycerides, a common form of fat in the blood, rose steadily, for three to four hours and heardly budged. It turns out that those individuals who barely responded have a mutation that disables one of their two copies of a gene called apoC-III. The gene codes for a protein, APOC3 that normally slows the breakdown of triglycerides.” What epidemiologic phenomenon is described by the investigators’ findings?
- Bias
- Conofounding
- Effect modification
- External validity
25. Prevalence effects which of the following measures of a screening program?
- Sensitivity
- Specificity
- Positive predictive value
- Negative predictive value
26. A study was conducted to evalue the association between TV viewing habits and attention deficit disorder. The investigator’s simple was restricted to children wo were home-schooled and ate a vegetarian diet. A reviewer questioned the study’s validity citing the extremely selective population studied (home-schooled, vegetarian children). Which type of validity does this criticism address?
- Internal validity
- External validity
- Recall bias
- Interviewer bias
27. A (hypotetical) cohort study was conducted by USF epidemiologists to assess the association between sleep deprivation (defined as <6 hours of sleep per night on average) and overall por academic performance (defined as a GPA <2.5 on a 4.0 scale). The investigators followed 3,000 freshman students over their four years at USF. Of 1,200 sleep-deprived students, 245 had GPAs <2.5, while 679 of the remaining students had a GPA of 2.5 or greater. The study was conduced with 85% statistical power. The 95% confidence interval for the risk ratio for por academic performance associated with sleep deprivation ranged from 0.26-0.89. Given the information about the 95% CI from the study on the association between sleep deprivation and poor academic performance, the p-value for the main association (wtih 95% CI 0.26-0.89) must be:
- Exactly equal to 0.05
- Less tan 0.05
- Greather than 0.05
- Insufficient information is given to conclude what the p-value might be.
28. The circumstance in wich different criteria are used to enroll study participants is known as:
- Random selection strategy
- Confounding
- Selection bias
- External validity bias
29. A study evaluated the use of a new cognitive test as creening tool for Alzheimer’s disease. As part of a case-control study, the results of 216 patients with definite or probable diagnosis of Alzheimer’s disease were compared to 432 controls without cognitive impairnent. The results are presented in the below.
True Disease
Status
+ –
Screening + 172 65
Test – 44 367
30-What is the specificity of the screening test?
- 72.6%
- 79.6%
- 85.0%
- 89.3%





