0314 - Confounding - EG Flashcards
Compare internal and external validity
Internal validity is if results reflect the true situation in a sample, external validity is if the results can be generalised to a broader population of interest.
Define selection bias. List some methods for how can it be controlled?
where the sample is not representative of the population of interest, therefore no generalisations can be made.Controlled by random selection; clearly defined: case-definition, eligibility (inclusion/exclusion) criteria, target population; strategies for high participation and retention of participants.
What is more problematic - random or systematic error?
Random error (imprecise) if true should cancel out and as such not overly effect results, however systematic error (inaccurate) can under or over estimate the true population.
What are methods that can help control measurement bias?
Clearly defined case definition; quality of measurement instruments: calibration, recall bias, interview tactics.
What is confounding?
The distortion of the effect between exposure and outcome due to the association of the exposure with other factors that influences the outcome.
What is an effect modifier?
An association between exposure and outcome due to a third variable. I.e season and risk of gastroenteritis
How does one differentiate between a confounder and effect modification?
strata the results, if there are SIMILAR RR between the strata then there is NO effect modification (but may be confounder if crude RR and adj RR differ by 10%); if there strata have DIFFERENT RR then there IS effect modification.
