Exposure measurement errors if not independent of other errors lead to study bias. Use of a 2 x 2 table, looking at dichotomous variable, avoids some of this non differential misclassification and reduces bias

Greenland S; Gustafson P

Exposure measurement errors if not independent of other errors lead to study bias. Use of a 2 x 2 table, looking at dichotomous variable, avoids some of this non differential misclassification and reduces bias

Greenland S, Gustafson P2006Accounting for independent nondifferential misclassification does not increase certainty that observed association is in the correct direction. Am J Epidemiol

Summary

No abstract is available for this paper. It was published in Accounting for independent nondifferential misclassification does not increase certainty that observed association is in the correct direction. Am J Epidemiol in 2006 by Greenland S and colleagues.

Part of CIRS Nexus · Curated by Lauren Lowenstein, The Biotoxin Lady