Friday, January 29, 2010

Big 5 personality traits and IQ, voting behavior at the state level

The Big 5 personality traits are intriguing, but aggregate measures are often unsatisfactory. I've previously posted on the counterintuitive inverse relationship between credit scores and conscientiousness at the state level as an illustration of this. Continuing that approach, the correlations with voting and estimated IQ for each of the five factors, as measured by Jason Rentfrow et al, follow*:

Voting for McCain and...r-valuep-value
Extraversion.10.50
Agreeableness.28.05
Conscientiousness.41.00
Neuroticism(.17).23
Openness(.46).00

Estimated avg IQ and...r-valuep-value
Extraversion.01.93
Agreeableness.05.70
Conscientiousness(.22).12
Neuroticism(.04).76
Openness(.02).87

Because the data for all variables are by state ranking rather than by specific numerical value, linear correlations (for which I'm measuring) are likely to appear more robust than they actually are.

If a positive relationship between openness and voting for McCain was revealed, I'd really be ready to throw in the towel on inter-population comparisons, for reasons identified by Steve Sailer:

Personality testing really needs some way to norm across subcultures. It seems like it does a fine job on, say, distinguishing among University of Illinois psychology majors, but once you get outside of a particular group with the same references, it falls apart on the between-group predictions (while, apparently, remaining okay within group).
However, the two correlate in the expected way, so it doesn't appear we're trudging aimlessly through bitumen.

The slight positive correlation between agreeableness and conservative voting behavior doesn't strike me as surprising. Leftists seem to be more favorably inclined toward making their cultural and political opinions known than conservatives are, whether those opinions be solicited or not. Per capita, leftist causes also seem to draw more activists out to protest than conservative ones do--think gay rights demonstrations versus Nixon's Silent Majority. However, it contrasts with a 2006 study that found agreeableness, openness, and neuroticism correlated with voting for Kerry in 2004, while higher conscientiousness and extraversion correlated with voting for Bush.

The positive relationship between conscientiousness and supporting McCain is probably more expected. The attributes defining high conscientiousness tend to be celebrated as virtues by the Popular Right. They include being prepared, fulfilling duties and promises made, favoring structured settings to organic ones, being meticulous in one's work, etc.

The only relationship between personality traits and intelligence I've repeatedly heard or read about is the modest but positive relationship between openness and IQ. That is not evident here, nor are statistically significant correlations of intelligence with the other four traits.

* R-values show the strength of the correlation between variables and range from -1 to 1. Negative numbers indicate an inverse relationship (as one goes up, the other goes down), positive values show a positive relationship (as one goes up, so does the other), and zero indicates no relationship whatsoever. For our purposes, p-values essentially give the probability that the relationship is meaningless. For those of .10 or more, correlations should taken with little more than a grain of salt.

7 comments:

Anonymous said...

"The only relationship between personality traits and intelligence I've repeatedly heard or read about is the modest but positive relationship between openness and IQ."

Psychologist David Nettle points out that this relationship is because there are questions for openness that will cause a high IQ person answer in such a way that the two will correlate. E.g., "I enjoy complexity". He claims this allows openness to become "infected" with high IQ in an artificial way. One wonders if the creators of the Big 5 set things up that way knowing it would make liberals appear more intelligent.

Audacious Epigone said...

Openness as a way of intellectual posturing, then? Interesting. That's the reaction I often have hearing SWPL-types explain how open they are to new ideas and experiences.

Beej said...

A slight variation to the big five personality traits is disucssed by Jonathan Haidt. If found it quite interesting. There is a bit of "poking fun" in the early part, but please watch up to the first 4 minutes where the speaker starts to discuss how libers and conservatives can learn how the other thinks and begin to learn how to work better together.

http://www.ted.com/talks/jonathan_haidt_on_the_moral_mind.html

Given the high correlation between liberal thinkers and science professionals, I would expect to see a relationship between IQ and generally liberal traits. I doubt the creators of the big 5 intended to make liberals seem more intelligent, but insted recognized this...


People with a liberal bent tend to be free thinkers, but also do not tend to do as well in a highly structured environment. Both perspectives are necessary.

rob said...

It's nteresting that only Openness correlates with IQ. In this paper Rushton et al constructs a general factor of personality, which "underlies" reasonably large chunks of OCEAN (negatively with neuroticism).

The GFP doesn't correlate with IQ(IIRC), which implies that the "part" of Openness that correlates with IQ is doesn't correlate with general personality.

In general, personality traits are part of how the mind works. I don't think there are many high-level mental functions that don't correlate with g.

Beej, while openness correlates with IQ and also correlates with liberalism, there does not mean that liberalism correlates with IQ. After all, Agreeableness correlates with Openness, but it does not correlate with IQ.

One must also ask, openness to what? Liberals are more open to government regulation, but less open to free markets. More open to environmental explantions of human variation, and quite a bit less open to hereditary explanations... Perhaps people are right to distinguish liberals from Democrats.

Anonymous said...

"Openness as a way of intellectual posturing, then? Interesting."

Well openness being diagnosed via Big 5 in such a way that it naturally captures IQ. "I enjoy complexity" is more of an IQ question than an openness question, since a close-minded person with a high IQ might love himself some complex high level maths.

But if we - the liberal creators of the Big 5(an assumption on my part) - can make openness, which corresponds w/ liberalism, also correspond w/ IQ, then poof! Liberals come out smarter than conservatives.



Hey, AE, you mentioned awhile ago making a post about G Miller's book "Spent". I think you had some criticisms.

Get on that broham!

Audacious Epigone said...

Beej,

Yes, I remember watching that TED talk awhile back. Haidt's work seemed interesting to me, though I've not read any of his books on the political dimensions of personality traits. Should I?

Anon,

Miller answered many of the questions I'd felt needed answering in the latter parts of the book (I believe I was referring to The Mating Mind, not Spent, though). At its completion, I felt like I didn't really have much to say after all. Sorry for coming up short.

Anonymous said...

"Openness to Ideas" and "Openness to Values" are theorised by some to provide a separate factor, Intellect, from "Openness to Fantasy/Aesthetics/Actions/Feelings" which form an Openness factor.

Supposedly only scores on the Intellect factor/subsets show correlations with brain performance and intelligence (http://www.ncbi.nlm.nih.gov/pubmed/19857008), but the Openness and Intellect subsets seem to show a mutual correlation in an unsorted population. Apparently the general correlation holds true for Liberals and supposedly for Conservatives (it might be nice to imagine that Liberals were all people with high openness to fluffy stuff but not actual ideas or the inverse for Conservatives, but that doesn't seem to be the case).

I expect the "closed minded" person who enjoys complex maths or philosophy would probably score highly on the Intellect subset but not the Openness subset. Likewise we can all imagine people who are very open to fantasies and feelings and aesthetics and actions but are as dumb as a box of rocks.