correct for

[loveyh]

Similarly, researchers in the famous Framingham study identified a significant correlation between coffee drinking and coronary heart disease. However, most of this correlation disappeared once researchers corrected for the fact that many coffee drinkers also smoke. If the confounding facotrs are known, it is often possible to correct for them. However, it they are unknown, they will undermine the reliability of the causal conclusions we draw from epidemiological surveys. (extracted from "Brute science: Dilemmas of Animal Experimentation)

Q: In the above sentences, what does "correct for" exactly mean?

 

[owlman5]

It should mean something like to eliminate confounding factors from a study/analysis.

 

[lingobingo]

I take it to mean that research findings can be corrected (adjusted) for (to take account of) whatever “confounding factors”, or confounding variables, are thought to be giving a false result.