KEY Review Questions for Chapter 9
- (1) The population to which the effect
applies (2) The exposure ("independent
variable") (3) the health outcome or disease
being studied ("dependent
variable") (4) the expected dose-response (5) the expected time-response
and (6) extraneous factors that could
confound results or modify the effect of the exposure
- This is the description of how the exposure dose influences the risk of the outcome.
- This is a description of the temporal relationship between the exposure
and the occurrence of the outcome.
- False. For an excellent commentary, see Maclure, M. (1991). Taxonomic axes of epidemiologic study designs: a refutationist perspective. Journal of Clinical Epidemiology, 44(10), 1045-1053
[Link
to abstract].
- In experimental studies, the exposure is assigned according by mechanisms built into the
study protocol.
In non-experimental studies, the exposure is studied without interference.
- Randomization balances groups with respect to measured and unmeasured
extraneous factors that could otherwise confound results.
- This means that you should design the study to try to isoloate the effects
of the exposure.
- This is an aggregate-level measurement
- This is a cross-sectional measurement
- True
- Cohort studies evaluate the experience of all people in a source population.
Case-control studies evaluate the experience of all cases and a
subset
of non-cases.
- Ecological studies use aggregate-level data.
- Longitudinal measurements require follow up of individuals over
time. Cross-sectional measurements related to a single cross-section in
time.
- (1) = Ecological
(2) = Cross-sectional study
(3) = Case-control
(4) = Cohort
(5) = Trial
Last update: 11/23/2008