Final Exam 3 Flashcards
What are observational pros
High sample power, exclusion/inclusion are relaxed causing increased external validity, measure long-term outcomes and detect more rare events, completed quickly
What are observational cons
Biases cannot be easily removed, observational trials are subject to missing or incorrect data, little control over the variable, can only establish associations (no causality/ correlation does not mean causation)
What are common types of internal validity flaws in observational trials
Selection bias, confounding bias, information bias
What is selction bias, what are the types of selcetion bias
The method of selecting subjects leads to a result that does not reflect the general population/ sampling bias, detection bias, response bias
What is confounding
infuential risk factor/exposure that is unequally distributed between groups leading to incorrect conclusions that the actual exposure/risk factor is associate with the outcome
How does selection bias become managed
Use objective criteria to identify cases and controls, inclusion/exclusion criteria must be a priority, need high participation rates to better validate survey data
What are ways that confounding is controlled
Create inclusion/exclusion criteria to minimize variation (improve internal validity), matching: selecting a control for every case that is very similar, stratification: create subgroups, multivariable analysis: regression analysis that will inherently adjust for variances
How can information bias be managed
use a prospective cohort design, observing and recording bias, record the reasons for loss to follow-up, observational trials have higher external validity