Exposure misclassification bias can profoundly affect results of an observational study on the impact of an exposure on health.


Misclassification errors may include:

(1) a person with past exposure classified as never exposed

(2) a person with an exposure who denies the exposure

(3) a person without an exposure who is said to have been exposed

(4) overestimation of exposure (greater than it actually was)

(5) underestimation of exposure (less than it actually was)


Misclassification errors can arise from:

(1) insufficient resources to perform an adequate study

(2) failure to objectively check data

(3) being overzealous

(4) lying


The impact of misclassification can be affected by:

(1) distribution of errors (one-sided vs offsetting)

(2) percentage of subjects affected

(3) probability of the outcome


To read more or access our algorithms and calculators, please log in or register.