Atheists are about 0.5σ smarter than people who find the Jeebus torture-porn cult convincing. They (atheists) test roughly in line with the actual result of secularised US Jews (i.e., a mean of ~107 – nobody worth spit still believes the nonsense claim of μ = 115 for Ashkenazi IQ).
The stereotype about atheists being pricks notwithstanding–or, to put it in psychological terms, low on the personality trait of agreeableness!–the research clearly backs his assertion regarding the correlation between religiosity and measured intelligence.
The GSS shows a similar result. Mean IQ by theistic orientation converted from Wordsum scores assuming a mean IQ of 100 and a standard deviation of 15 follow. To avoid racial confounding, results are restricted to non-Hispanic whites:
Excerpting from an abstract of a Lynn, Nyborg, and Harvey paper, Kratoklastes also notes:
Evidence is reviewed pointing to a negative relationship between intelligence and religious belief in the United States and Europe.
The phrase “religious belief”, rather than a phrase like “religious observance” is important. Mean IQ by frequency of religious worship attendance:
Among firm theists, as attendance frequency increases, so does intelligence. The mean IQ among firm theists who never attend worship services is 96.1. Among firm theists who attend at least weekly, it is 100.1. If God is real and He wants you to worship Him, the smart sheep will be in the pews on Sunday morning while the dumb goats sleep in.
Among non-believers, the trend tends modestly in the other direction (mean IQ of 104.5 among atheists and agnostics who never attend; mean IQ of 99.1 among those who attend at least weekly).
The highest mean IQ as measured among the categories discussed, at 108.1, is found among agnostics who attend worship services less than monthly (Christmas eggs).
GSS variables used: GOD(1)(2)(3-5)(6), RACECEN1(1), HISPANIC(1), WORDSUM, BORN(1), ATTEND(0)(1-3)(4-6)(7-8)