Thursday, September 29, 2005
A Statistical Update
Yesterday we looked at the very specious reporting concerning the apparent coincidence of religiosity and poor crime and social factors in the United States as compared to less religious european countries.
Thanks to commenter Rebecca I can point you to this great post from statistician/blogger Magic Statistics. He, like I, points to the confusion between correlation and causation, but goes on to point out some methodical errors in the study, including
Thanks to commenter Rebecca I can point you to this great post from statistician/blogger Magic Statistics. He, like I, points to the confusion between correlation and causation, but goes on to point out some methodical errors in the study, including
- an equation of rejection of evolution with religious faith
- no explanation of what countries are chosen for study, and what are not
- no explanation for the selection of times frame for the study
- no examination of other possible correlative factors
The line between "study" and hit piece is pretty fine, but I think it is fair to say we have crossed it here.
