EBVM Flashcards
Numerical variables
variables which are quantitative
Categoric Variables
variables that fit into categories, such as male or female (which is also binary)
selection bias
before the study begins, when selecting the subjects for the study
EG: choice of comparison groups, non-response bias, missing data, loss of follow up, healthy worker effect
confounding bias
when two factors are linked to each other and the outcome, one is inaccurate
EG: coffee drinking-x-> lung cancer, smoking—> coffee drinking, smoking—> lung cancer
misclassification bias
Also known as measurement bias for continuous variables
incorrect classification of outcome or exposure
imperfect sensitivity or specificity
cross-sectional study
snapshot of info about exposures and disease
what does a cross-sectional study measure?
Point prevalence, relative risk, attributable risk
limitations of a cross-sectional study?
prevalence is an outcome- can’t differentiate between factors for persistence and development of outcome
exposure and outcome are measured at the same time, can;’t differentiate cause and effect
Cohort study
follow a target group for a period of time and compare the outcomes in exposed to unexposed
what is the unit of study for a cohort study?
animal time
why would an animal be removed from a cohort study?
death, sale, disease (not the one being investigated), no longer in the risk category (no longer lactating)
what can a cohort study measure?
incidence rate, relative risk, attributable risk, attributable fraction
advantages of a cohort study?
- can study several diseases simultaneously,
- obtain an estimate of disease incidence,
- temporal relationship between exposure and outcome- inferring causality
Disadvantages of a cohort study
large study population takes a long time, costs lots
