Liberals Are Neurotic and Conservatives Are Antisocial

An awful lot of cyberink has been spilled over the liberal-conservative EEG study published online in Nature Neuroscience a few days ago. The Neurocritic is not an expert in social psychology or personality, so to be fair in the sequel to The Error of Prognosticating Political View by Brain Wave, I decided to read [skim] the 2003 paper by Jost and colleagues, Political Conservatism as Motivated Social Cognition (PDF here). An obvious question for the uninitiated is whether there is any truth to the title of this post, other than the inference I made based on the error-related negativity literature: ¹

Liberals showed larger ERN waves than conservatives when mistakenly responding on No-Go trials. However, so do individuals with clinical diagnoses such as obsessive-compulsive disorder (Gehring et al., 2000) or major depressive disorder (Chiu & Deldin, 2007).

On the other hand, individuals with schizophrenia (Mathalon et al., 2002) or psychopathy (Munro et al., 2007) show smaller ERN waves than control participants. These findings extend to the normal population, i.e., people who do not fit the criteria for a clinical diagnosis, but who score higher or lower on certain traits. For example, people who score high on negative affect have bigger ERNs (Hajcak et al., 2004), while individuals with "externalizing psychopathology" have smaller ERNs (Hall et al., 2007). Does this mean that liberals are neurotic and conservatives are antisocial? Since these were not assessed along with political orientation, we can only hazard a guess...

Actually, we may not need to hazard a guess, because such measures may have been already obtained from these subjects, as stated in the Supplementary Methods:

A measure of political attitudes was embedded in a larger set of personality and attitudes surveys completed at the every [sic] beginning of the experimental session.

So a paper about neurotic liberals with large ERNs and antisocial conservatives with small ERNs may yet appear in a journal near you. This begs the question,

Perhaps there were any other correlations? I wonder what the odds were of a stastically [sic] significant finding given the number of personality and attitude surveys.

But back to the 2003 paper. I do remember that it raised a big stink in the media at the time, not surprising given a press release like this:

Disparate conservatives share a resistance to change and acceptance of inequality, the authors said. Hitler, Mussolini, and former President Ronald Reagan were individuals, but all were right-wing conservatives because they preached a return to an idealized past and condoned inequality in some form. Talk host Rush Limbaugh can be described the same way, the authors commented in a published reply to the article.

Here's the abstract of the paper (Jost et al., 2003), which really is all about conservatism: ²

Analyzing political conservatism as motivated social cognition integrates theories of personality (authoritarianism, dogmatism–intolerance of ambiguity), epistemic and existential needs (for closure, regulatory focus, terror management), and ideological rationalization (social dominance, system justification). A meta-analysis (88 samples, 12 countries, 22,818 cases) confirms that several psychological variables predict political conservatism: death anxiety (weighted mean r= .50); system instability (.47); dogmatism–intolerance of ambiguity (.34); openness to experience (–.32); uncertainty tolerance (–.27); needs for order, structure, and closure (.26); integrative complexity (–.20); fear of threat and loss (.18); and self-esteem (–.09). The core ideology of conservatism stresses resistance to change and justification of inequality and is motivated by needs that vary situationally and dispositionally to manage uncertainty and threat.

OK then, "resistance to change" appears to extend to a lowered ability to inhibit habitual responding and a diminished brain response to such mistakes (Amodio et al., 2007). It's still not clear to me, however, how brain systems responsible for personality constructs like intolerance of ambiguity and needs for order, structure, and closure overlap with brain systems responsible for not pressing a key to a letter presented 20% of the time.

Now back to the 2007 NN article again. I hate to belabor the point, but it's only fair to examine the paper on its merits, not on the ideologies of the authors, no matter how much I might agree with them. But really, this blog comment sums it up:

It is useless to try to infer anything from this study. The experimenter has 26 liberals and 7 conservatives. Not exactly equal sample sizes, and not a big enough sample of conservatives to gain any inference about the range of responses here, even if you buy into the non-sequiter introduction set out by the authors. [NOTE: the following objections were addressed by the first author.] A thought experiment: would the correlation hold up if the two biggest outliers were excluded. Doesn't look like it. No mention of possible sex differences etc. usually discussed in these papers. The authors fail to address these confounds even in the supplementary info. A sad commentary on the state of science when a journal with this level of impact publishes a study of this low quality.

Posted by: John Broussard | September 11, 2007 6:48 PM

Huh. So this would imply that the brain wave data shown in Fig 1b of Amodio et al. (2007) were averages from 7 conservatives and ~26 liberals!

And a very important bit of data -- mean accuracy rates for conservatives and liberals -- didn't appear anywhere in the paper [someone correct me if I'm wrong], but was published in this newspaper article:

In the 400 easy trials, just about everyone got it right.

But in the 100 tough trials, when students saw the letter that meant they shouldn't press a button, self-described conservatives pressed the button anyway nearly half the time - an error rate of 44 out of 100.

Liberals fumbled about a third of the time, with an error rate of 34 out of 100.

Neurophilosophy reported slightly different figures [not sure where he got them]:

It was found that those who considered themselves to be conservatives made more response errors when upon presentatin of the infrequent letters than those who considered themselves as liberals (respectively, 47% and 37% of the time).

Have any of you ever published papers in less prestigious journals without including mean accuracy and reaction time values for your comparison populations?? There was absolutely no information about RTs at all, so we don't know whether there was a speed-accuracy trade-off in the conservatives (a "reckless" and disinhibited response style) or whether they were "conscientious" (RT comparable to [or slower than] liberals), but just couldn't stop themselves from pressing the key on No-Go trials.

Enough already with the ranting.

Peace. Shana Tova.


Footnotes

¹ See previous post for a list of those references.

² This paper doesn't discuss the psychological variables that predict liberalism, so in the end we don't know if liberals are neurotic; the incidence of antisocial traits in conservatives was not reported.

References

Amodio DM, Jost JT, Master SL, Yee CM. (2007). Neurocognitive correlates of liberalism and conservatism. Nature Neurosci. Published online 9 September 2007.

Jost JT, Glaser J, Kruglanski AW, Sulloway FJ. (2003). Political conservatism as motivated social cognition. Psychol Bull. 129:339-75.