A Good Read
There’s a certain cargo cult associated with these sorts of correlations.
If there are more than 3 or 4 variables that have an influence on something, we very quickly lose our ability to say anything meaningful about the situation. It’s true in health, the environment, and it’s also true in Education.
Knowing that the technique is terribly flawed and asking yourselfâwhich you shouldnât have to do because you ought to be told by the journalist what generated these dataâif the study is subject to self-selection effects or confounded variable effects, and if it is, you should probably ignore them. What I most want to do is blow the whistle on this and stop scientists from doing this kind of thing. As I say, many of the very best social psychologists donât understand this point.
Source: The Crusade Against Multiple Regression Analysis | Edge.org