Behavioral (Epi)

Question 1

Compare confounding and effect modification.

Answer 1

Confounding – a variable that is not controlled for has an effect on the outcome
Effect modification – a variable mediates the effect of an intervention and is controlled for through stratification (not a bias)

Question 2

What are α, β, and 1 - β?

Answer 2

alpha = maximum probability of making a type 1 error the researcher is willing to accept (usually 0.05)
beta = probability of committing a type 2 error
1 - β (1 minus β) = probability of rejecting a null hypothesis when it is truly false, i.e. statistical power (to find a true relationship if it exists)

Question 3

What is capitation? What does it incentivize?

Answer 3

when physicians get paid a set amount for the number of patients they care for; this incentivizes preventive health

Question 4

How do you calculate NNT?

Answer 4

1/ARR

Question 5

How do you calculate ARR (absolute risk reduction)?

Answer 5

event rate in placebo group - event rate in control group

Question 6

What does the power of a study depend on?

Answer 6

sample size

Question 7

What is the formula for odds ratio?

Answer 7

ad/bc

Question 8

What is the formula for relative risk?

Answer 8

(a/a+b)/(c/c+d)

Question 9

What measure is used to assess test reliability?

Answer 9

Coefficient of variation (CV) - calculate by dividing standard deviation of a set of repeated measurements by the mean, and express as percent; can be used to compare precision of tests with different units of measurement and normals

Question 10

How can you tell the difference b/w confounding and effect modification using stratification?

Answer 10

If you stratify by a confounding variable, there will be an insignificant difference b/w the groups; if you stratify by an effect modifier, the groups will significantly differ