My article on IQ and religious belief

[bpesta22]

Greetings.

I just got an article accepted today on IQ and religious belief. I thought some here might be interested in it, given our many debates about IQ.

It's in press in the Elsevier journal, Intelligence.

I own the copyright to the pre-print version which is available here:

http://ourworld.cs.com/bpesta22/religion_IQ_bertsch_pesta.pdf

If you do read it, thanks, and criticisms are indeed welcome.

[ImaginalDisc]

Before I read it, I suggest you convert that a .pdf. If you don't have access to Adobe Acrobat professional or a similar program, I use the OpenOffice suite, which is compatible with Microsoft and lets you turn any document into a .pdf with one button.

The simple correlations offer only moderate support for Hypotheses 1 and 2. While in the predicted direction, only half the correlations emerged as significant. Namely, higher IQ and ECT performance were associated with less belief in the sole truth of one’s religion. Higher ECT performers were also more likely to question their religious beliefs.

The resulting WPT correlations were
-.33, -.14, and .14, for sectarianism, scriptural acceptance, and religious questioning, respectively. Even after correction, the correlations offer only partial support for Hypotheses 1 and 2.

The use of sectarianism scores successfully classified 65% of the cases in each of the two lowest IQ groups, 43% in the third IQ quartile, and 75% in the highest IQ group. The model could not be improved by the addition of any other variables. The same analysis using religious questioning was only marginally significant (p = .06), and correctly categorized 52% of the lowest IQ quartile scores, 64% of the second, 47% of the third and 65% of the fourth.

Results for scriptural acceptance were less clear, as differences were not monotonic across the IQ and ECT quartile groups.
Emphasis added in all cases.


Not encouraging results, but not surprising. Smart people can cleave to stupid ideas, too.

At a practical level, important social issues are often decided as much by religious beliefs as by reason. Examples include whether to ban gay marriage and stem cell research, or whether creationism should be taught alongside evolution by natural selection in science classes. Discovering a negative relationship between measures of general mental ability and literalist religious beliefs could partially influence the amount of weight given to arguments generated by faith- versus reason-based criteria.

No, the biggest factor for deciding on how much weight religious or faith based ideas should have is whether or not they're true. Even if such ideas were the domain of the fantastically intelligent, they could be dismissed just as easily because those specific ideas are hogwash. You try telling the electorate that religiously motivated bigotry should be ignored because only stupid people think that way. It's not true, it's not politically viable, and it's not even relevant.

[slingblade]

ID, and B, I have to ask if that last quoted bit is BPesta's prediction of how such information could be used, in the public sector, as a barrier or discrimination to people, or that he thinks this is what the numbers reflect?

I mean, such a correlation could be used to influence others, and having a study to confirm it adds great weight. I agree that the deciding factor ought to be the validity of the argument, not the intellect of the argument-holders. But if it's a factor, and it seems to be, someone somewhere will try to use it, somehow.

I'm just not clear, because it could be taken either way (or a third way I'm not yet seeing). So.

[CFLarsen]

And there we were, being told that individual IQ scores didn't matter - it was the IQ of a specific population.

How things change.

[bpesta22]

Before I read it, I suggest you convert that a .pdf. If you don't have access to Adobe Acrobat professional or a similar program, I use the OpenOffice suite, which is compatible with Microsoft and lets you turn any document into a .pdf with one button.





Emphasis added in all cases.


Not encouraging results, but not surprising. Smart people can cleave to stupid ideas, too.



No, the biggest factor for deciding on how much weight religious or faith based ideas should have is whether or not they're true. Even if such ideas were the domain of the fantastically intelligent, they could be dismissed just as easily because those specific ideas are hogwash. You try telling the electorate that religiously motivated bigotry should be ignored because only stupid people think that way. It's not true, it's not politically viable, and it's not even relevant.





ID: Thanks; some of the results were indeed weak and non-significant. You listed em all. We did find fairly consistent and moderately strong effects across groups separated into quartiles. I think one reason why it got accepted is that it's not obvious why judging line lengths and selecting letters should predict how strongly one believes in his/her religion.

I'm not able to PDF files. If you can do it easy, send me it and I will link to it rather than the word file (no big deal if you can't do this).

[bpesta22]

ID, and B, I have to ask if that last quoted bit is BPesta's prediction of how such information could be used, in the public sector, as a barrier or discrimination to people, or that he thinks this is what the numbers reflect?

I mean, such a correlation could be used to influence others, and having a study to confirm it adds great weight. I agree that the deciding factor ought to be the validity of the argument, not the intellect of the argument-holders. But if it's a factor, and it seems to be, someone somewhere will try to use it, somehow.

I'm just not clear, because it could be taken either way (or a third way I'm not yet seeing). So.

We spent a fair amount of time on that last paragraph. I guess what we wanted to say is how you interpreted it. We see this as one factor-- not the biggest factor; not the only factor-- that might be considered when weighing opinions created by fundamentalism.

Whether in practice anyone will do that (because of the politics issues raised by ID) is another story.

[bpesta22]

And there we were, being told that individual IQ scores didn't matter - it was the IQ of a specific population.

How things change.

Claus, you're right to call me on this. It's an aggregation thing. We're not advocating testing a person's IQ then weighting their opinions accordingly.

But when studying group differences, obviously, groups are the sum of individual scores (how else to compare two groups but to collect individual scores within groups).

Knowing an individual's IQ we could make a weak (but better than chance) prediction about how literally he/she might believe. It's an inductive argument, though, and given the weak effects, no one should be surprised if the prediction fails.

When we've aggregated things and are now making predictions about how groups of people might believe, the accuracy would be much better (though the inference would apply only to the group as a whole and not any individual in the group).

A single person's IQ probably won't tell you much about that person. But, get lots of people with the same IQ score and you can make very accurate inferences about how that group might differ from other groups with different IQs.

[bpesta22]

Thanks to Darat for PDF-ifying the article. Here's the new link. I wonder if the OP could be edited to change the link?

Thanks again

http://ourworld.cs.com/bpesta22/religion_IQ_bertsch_pesta.pdf

[leon_heller]

I just converted it and uploaded it to my web site. You beat me by a few seconds.

Leon

[bpesta22]

Thanks Leon.

I'm not sure it can be hosted on any site but my own for CR purposes. I'm cool with people linking to it from anywhere, but perhaps the document itself should not be hosted on other sites?

[Darat]

Changed your link to the PDF document.

[CFLarsen]

Swap "IQ" with "horoscopes". The argument is exactly the same:

A single person's horoscope does not tell us much about that person being e.g. a Leo or a Virgo. Individually, Leos can differ a lot. But, get lots people with the same sign, and you can make very accurate inferences about how Leos might differ from other groups with different signs.

It's an aggregation thing.

[bpesta22]

Swap "IQ" with "horoscopes". The argument is exactly the same:

A single person's horoscope does not tell us much about that person being e.g. a Leo or a Virgo. Individually, Leos can differ a lot. But, get lots people with the same sign, and you can make very accurate inferences about how Leos might differ from other groups with different signs.

It's an aggregation thing.

But now you've framed it as an empirical question and you can test it. Aggregate horoscopes and they still don't predict bunk. Aggregate IQ and you often get the single best predictor of anything else.

[ImaginalDisc]

But now you've framed it as an empirical question and you can test it. Aggregate horoscopes and they still don't predict bunk. Aggregate IQ and you often get the single best predictor of anything else.

Did you just perform a study which found extremely weak correlations between specific mental tasks and religiosity?

[CFLarsen]

As with astrology, this "often", so casually thrown around, isn't quantified.

Aggregate horoscopes and they do predict something: Not due to any celestial influences, but because of self-fulfilling prophecies. If Leos are constantly told that they are born leaders, and are treated as such during upbringing, naturally they will have a higher chance of being leaders. Likewise, tell a group of Pisces children that they are likely to become drunk suicidals, and they will kill themselves in a drunken stupor more often than other groups.

When we aggregate, of course.

[Darat]

Actually studying astrology in this way may very well bring up some significant and real differences in the groups since the star signs are date determined and the date people are born may have some affect on their later lives (as I believe some studies have indeed shown). However the mistake would then be to swap this correlation for a causation.

[bpesta22]

Did you just perform a study which found extremely weak correlations between specific mental tasks and religiosity?

The effect sizes were moderate on most of the measures comparing the lowest and highest scoring IQ (and ECT groups).

That said, I'm not sure what point you are making here?

[ImaginalDisc]

The effect sizes were moderate on most of the measures comparing the lowest and highest scoring IQ (and ECT groups).

That said, I'm not sure what point you are making here?

Ok, I'll unpack that for you. You said this:

But now you've framed it as an empirical question and you can test it. Aggregate horoscopes and they still don't predict bunk. Aggregate IQ and you often get the single best predictor of anything else.

A, your data show very weak correlations. Correlation is not causation. You have thus far not shown that any individual task or aggragate of tasks causes anything, ergo, a person's score on any tasks do not predict anything you have evalutated with reasonable confidence.

B, "best predictor?" ************! You know what's a better predictor of a person's religion than their mental task scores? The religion of their parents. You know what's a better predictor of a person's academic success than their mental task scores? The socio-economic status of their parents.

[bpesta22]

Ok, I'll unpack that for you. You said this:



A, your data show very weak correlations. Correlation is not causation. You have thus far not shown that any individual task or aggragate of tasks causes anything, ergo, a person's score on any tasks do not predict anything you have evalutated with reasonable confidence.

B, "best predictor?" ************! You know what's a better predictor of a person's religion than their mental task scores? The religion of their parents. You know what's a better predictor of a person's academic success than their mental task scores? The socio-economic status of their parents.

Show me anywhere in the article where I assert causality (i.e., claim that religion causes you to be stupid).

You got any data to support your claims? I do. Here's state level AGGREGATE data showing the correlation between state IQ and various religious beliefs:

% of respondents w/in states that agree with this statement:
I am certain god exists (r = -.43)
Religion is very important (r= -.54)
I pray daily (r= -.59).
My prayers are answered regularly (r = -.64)
The bible is literally true (r = -.46)

State IQ correlates -.45 with the % of Evangelicals in the state.

-.70 with murder rates
.44 with college degrees (%)
-.71 with teenage moms (%)
-.70 with families below poverty
.39 with state income levels.

As a specific test of your hypothesis, the correlation between state religiosity and IQ is -.55. It's still .-31 with controlling for income and poverty levels.

Conversely, the correlation between income and poverty levels and IQ is .52. controlling for religion, it's no longer significant (.22).

Religious beliefs explain why IQ and poverty correlate.

Poverty does not explain why religious beliefs and IQ correlate.

That's aggregation.

My study would need to compare 100s of people with IQs of say 85 to 100s of others with IQs of 115 to show how aggregation strengthens results. Is there any doubt such a study would show huge differences in mean religiosity across the groups?

[CFLarsen]

I would have to answer "Yes" to "Religion is very important".

Sorry, but religion is very important. To other people, definitely, because so many are religious, and therefore it has a major influence on the world. But certainly also to me, since I spend a considerable amount of my time arguing against religion.

Does "state IQ" correlate with the % of Evangelicals in Denmark? It would have to, if we were to argue that "state IQ" correlates with the % of Evangelicals in the US states.

Likewise, does the Danish "state IQ" correlate with murder rates in Denmark? Gee, given the comparably much lower murder rate in Denmark, Danes must have a far higher IQ than Americans.

Hey, I can live with that!