Nyborg certainly doesn’t suffer from the handicap of objectivity. He also does not seem to know the religious affiliation map very well. He classifies as “Liberal”, Episcopal/Anglican, Jewish, Methodist, Presbyterian, Lutheran, and Protestant (Other). In fact, all of these orientations shade from ultra-liberal to very traditional - and the more traditional ones tend to have more teens and to hang on to them through youth clubs. Protestant (Other) are often stricter than your average church (which, in many cases, is how they came to be Other).

He classifies as “Dogmatic,” Disciples of Christ, United Church of Christ, Roman Catholic, Mormon, Bible Church, 'Muslem' (sic), Holiness, Baptist, and Pentecostal. He is surely mistaken about the Disciples of Christ and United Church of Christ, and - as a congregational denomination - Baptists are all over the map. He is on firmer ground with Catholics, Mormons, and Muslims, in the sense that they have non-negotiable core beliefs. On the other hand, sporadically observant Catholics are very numerous, and it’s not clear in what sense they are dogmatic. Overall, Nyborg creates the distinct impression of not having the depth of background needed for a more informative classification.

That last point, about sporadically observant Catholics, is quite important when one considers his methodology: Manchester University physicist David Tyler had a look at Nyborg’s survey design and commented that it “needs scrutiny.”

For example, when comparing IQs of atheists, agnostics and religious people, he selects sample populations in proportion to demographic data. It means that he uses 985 Roman Catholics, 541 Baptists, many hundreds of other categories, but only 103 agnostics and 39 atheists. Since the latter two groups are critical to the outcome, the small sample size does appear to be suspect.

Put another way: If half of the Catholics and Baptist teens are sporadically observant and doctrinally indifferent (no unusual state of affairs), religious orthodoxy collapses as a predictor of IQ. So it is not clear just what Nyborg is measuring. Social class is a possibility.

Perhaps Nyborg’s paper’s most useful function is to demonstrate something quite other than what he intended: That one can show a direct or inverse correlation between any two characteristics, provided that the numbers are defined, selected and/or weighted to achieve that result.

It’s worth noticing that this study on teen intelligence and Lynn et al’s study on national intelligence feature similar problems: Lynn et al. ignored the impact of personal freedom. Nyborg failed to classify religious groups’ stances or member commitment in a meaningful way. As a result, Lynn succeeded in demonstrating only something he probably didn’t intend: That intellectual freedom correlates both with higher IQ and a variety of stances in relation to belief in God. Nyborg probably didn’t measure anything meaningful. But continuing research into the relationship between religious affiliation and social class/income might shed some light on tested IQ scores.

Next week: Religion doesn’t just rot your intelligence, it shrinks your brain!
Last week: Does religion rot your intelligence?

Denyse O'Leary is co-author of The Spiritual Brain.