The issue is whether or not the study controlled for such things. However, as it is a meta-analysis, we have to look at the actual studies to really know. And having started that process, we might realize that in this particular study, of the 63 analyzed studies about a 3rd were conducted in this century. About 1/6 were conducted before 1960, and most were conducted before 1980.
But the fun doesn't stop there! Because not all studies were equal. This is true in many ways, but a rather fundamental one is number of subjects. Most had between ~100 & ~200. But 3 studies had over a 10,000. What were they?
Why Liberals and Atheists Are More Intelligent
Deptula, D. P., Henry, D. B., Shoeny, M. E., & Slavick, J. T. (2006). Adolescent sexual behavior and attitudes: a costs and benefits approach. Journal of Adolescent Health, 38(1), 35-43.
[this one was great. They asked a bunch of kids ~15 years old about sex, and then for some controls threw in a 1-4 "how religious are you" question and administered one cognitive test. What was the most statistically significant indicator of intelligence? Age. Remarkably, scientists determined that being young is correlated with things like acting like idiots. Just about any parent of teenagers could have told them that, but science requires careful observation and logic, not basic common sense. It's too common, and that's the last thing scientists want to be.]
Sherkat, D. E. (2010). Religion and verbal ability. Social Science Research, 39(1), 2-13.
[this one was even better. 90% of the sample were religious. Why? Because the study was really testing whether or not more fundamentalists attitudes towards the bible predicted verbal ability, and thus wanted more Christians. I can take that data and show that being Jewish or Episcopalian is a better predictor of intelligence than being non-religious.]
Then there are the studies that I still can't figure out how they managed to work into their results, such as
Corey, S. M. (1940). Changes in the opinions of female students after one year at a university. The Journal of Social Psychology, 11(2), 341-351.
Many of the studies didn't focus on or actually include measures of intelligence and/or religiousity.
Wouldn't that make converts to a religion equally likely to be intelligent? And it would make agnostics the stupidest, because they don't have an opinion other than "I dunno"?
Intelligence is an important thing to study, although I think far too many studies have concentrated on using education (along with other variables) to predict intelligence rather than look into what the quality of education one receives can predict.
Because this is not by any means the most studied relationship between intelligence and something else:
"A BlackWhite group difference on intelligence test scores has persisted in the literature for over 90 years. Currently, the group IQ mean for Blacks (85) remains about one-standard deviation below the group IQ mean for Whites (100; see, e.g., Neisser et al., 1996, Lynn, 2006 and Rushton and Jensen, 2006). Though the difference exists, no consensus as to its cause is likely forthcoming. Some argue that research here is flawed because race-based classifications are invalid (see e.g., Sternberg et al., 2005 and Tate and Audette, 2001), or because a single, global IQ score cannot adequately represent human intelligence (see, e.g., Gardner 1983). Others argue that BlackWhite differences are realdue neither to cultural, nor test biasand at least partly driven by genes (see, e.g., Herrnstein and Murray, 1994, Rushton and Jensen, 2005 and Gottfredson, 2005a)." (emphases added)
Pesta, B. J., & Poznanski, P. J. (2008). BlackWhite differences on IQ and grades: The mediating role of elementary cognitive tasks. Intelligence, 36(4), 323-329.
A glad as I am that at the very least there are innumerable and much more obvious reasons for the racial disparity (and nobody is citing The Bell Curve as reliable literature), I do find it curious that when religiousity or political conservatism are used to predict intelligence, we never find the argument that the entire approach to intelligence testing is flawed as there are simply to many factors for "a single, global IQ score" to "adequately represent human intelligence".
Give me any cohort (students, randomly sampled and appropriately weighted individuals across the country, selective/controlled sampling from geographic regions, mosque/church/temple/Dawkins fans, etc.) and scores that approximate continuous variables like IQ scores or pre-college scores, and I can show you just about any group is smarter than another by asking the right questions and using likert scale variables (1-5 is standard, but 1-6 is better because it removes the option to go for the middle). Why? Because one set of scores can be changed ever so slightly or grouped in just the right way, and the difference in the correlations will shift dramatically. This is especially true if, as in several of the studies used by the meta-analysis, I use something like structural equation modelling (SEM) which is either an extremely powerful statistical tool or a way to get whatever results you want.
Last edited:
