Confounding and effect modification: their significance in medical research
Abstract
Confounding and effect modification, two very important statistical concepts that may occur both in the statistical modelling of epidemiological data and other medically related data are seldom considered during the analysis of data. This is rather unfortunate as this may introduce some degree of bias in the statistical interpretation following the analysis.
Confounding and effect modification arc a major cause for concern in medicine. Epidemiology is an example of that branch of medicine where confounding and effect modification arc commonly applied.