Saturday, September 20, 2008

Conservatives Are Neurotic and Liberals Are Antisocial

Since turnabout is fair play, we can no longer say that Liberals Are Neurotic and Conservatives Are Antisocial.

In fact, it's opposite day! But why?

It's because of a new Science article by Oxley et al. (2008). I'm sure you've seen the deluge of articles in the popular press saying that social conservatives1 are great big scaredy cats when it comes to loud noises and aversive pictures. The headlines are ranked in order of most to least sensationalistic:

Newsweek: Spiders, Maggots, Politics the conservative mind, illegal immigrants may =spiders = gay marriages = maggot-filled wounds = abortion rights = bloodied faces.
Wired Science: Conservatives Scare More Easily Than Liberals, Say Scientists
Deep-seated political differences aren't simply moral and intellectual: They're also biological.
BBC News: Political views 'all in the mind'
...people who are sensitive to fear or threat are likely to support a right wing agenda.
EurekAlert! (Rice University press release): Political attitudes are predicted by physiological traits
Is America's red-blue divide based on voters' physiology? A new paper in the journal Science, titled "Political Attitudes Are Predicted by Physiological Traits," explores the link.
Scientific American: Are you more likely to be politically left or right if you scare easily?
Here's a fun trick: scare someone you don't know, then guess whether they favor the death penalty and the war in Iraq based on how freaked out they got.2

People with stronger startle reactions are more likely to support ideologies associated with conservative American politics, including the Patriot Act, obedience and biblical truth — and less likely to favor gun control, foreign aid, abortion rights, gay marriage and pornography, according to research published in today's Science. Those who are slower to scare are more likely to harbor traditionally liberal politics.
NSF press release: Some Political Views May be Related to Physiology
People who react more strongly to bumps in the night, spiders on a human body or the sight of a shell-shocked victim are more likely to support public policies that emphasize protecting society over preserving individual privacy.
Before I add my $0.02 (which is worth even less now), I'd like to draw your attention to two critiques in the blogosphere. For a good discussion of behavior genetics and heritability issues, read Gene Expression on Conservatives Have More Fear:
By heritability, I mean that you can predict an X proportion of the variation on trait Y from the variation of genes. Does this mean that there was selection for conservatism and liberalism in the past? Does this mean that there are genes for conservatism and liberalism? NO!!! That's just plain retarded.
For another excellent critique, I highly recommend A spider on your face won't stop me from criticizing this article by John Hawks. He provides a list of 7 erroneous assumptions and methodological flaws:
1. Sampling: By going from a sample of over 1310 down to 46 individuals, the authors were in a position to cull the sample in many ways that might influence their outcome. ...
That was my number one question as well. There is nothing in the main paper or the supporting online material (PDF) about whether the authors excluded (or included) subjects with things like anxiety disorders, for instance. How they settled on the final n=46 is murky.
2. Presentation of results: The methods employed are multiple regressions, using the dependent as a continuous variable. Yet the graphs in the paper show the dependent as a dichotomous variable. The graphs are designed to show the result to be as visually large as possible, but in so doing, they don't report the actual results, which are visually much weaker (while still statistically significant).
I won't go through the list point by point, but other objectionable aspects include lumping together the moderates and the conservatives, having a "made-up category" as the dependent variable ("has the look of a fishing expedition"), and not controlling for religious attitudes (!!).

The only thing I can add to the conversation is a brief discussion of the published literature on startle and personality. But first let's return to the beginning with the inversion of this post's title. In a famous study from last year, David Amodio and colleagues showed that an EEG wave related to making mistakes in a speeded reaction time task (called the error-related negativity or ERN) was larger in liberals than it was in conservatives. The authors also wanted to relate the ERN wave to conflict monitoring to conclude that liberals were more flexible in their thinking:
...greater liberalism was associated with stronger conflict-related anterior cingulate activity, suggesting greater neurocognitive sensitivity to cues for altering a habitual response pattern.
OK, then. After questioning whether the ERN was even related to conflict monitoring, I reviewed the literature on ERN and psychopathology.3 Briefly, liberals and individuals with clinical diagnoses such as obsessive-compulsive disorder or major depressive disorder have larger ERN waves, while conservatives and individuals with schizophrenia or psychopathy 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, while individuals with "externalizing psychopathology" have smaller ERNs. Does this mean that liberals are neurotic and conservatives are antisocial?
Well, no, if you look at startle response, it's the reverse. Although the difference was not statistically significant,4 conservatives high supporters of socially protective policies had a larger eyeblink startle response to random loud bursts of white noise than liberals.

Fig. 3 (Oxley et al., 2008). Mean blink amplitude in response to all seven startling noises for high supporters and low supporters of socially protective politics. Bars are mean blink amplitudes (in millivolts). Difference of means tests for overall means: t = 1.64, P = 0.10.

But is the startle reflex comparable to the ERN brain wave? Hajcak and Foti (2008) say they're related:
We found that (a) the defensive startle response was larger following errors than following correct responses, and (b) the magnitude of the ERN predicted the degree of startle potentiation following errors.
And what about the skin conductance response to aversive vs. neutral pictures? The SCR was recorded in another experiment that showed the participants pictures of bunnies and bloody faces. The conservatives' SCRs to aversive images were larger than those in the liberals, but the two groups were comparable for the nonthreatening images. How is the SCR related to the ERN? Hajcak (et al., 2004) again:
An enhanced ERN has consistently been observed in anxious subjects and there is some suggestion that the ERN is related to general negative affective experience (NA). The ERN has been source localized to the anterior cingulate cortex-a structure implicated in the regulation of affective, response selection, and autonomic resources. Thus, the findings that autonomic measures and affective distress are related to response monitoring are consistent with anterior cingulate cortex function. ... Results indicate that high NA was associated with significantly greater ERN and error-related SCR...
The only way to reconcile the findings from Oxley et al. (2008) and Amodio et al. (2007) is to postulate that the conservatives in the latter study were unaware of making errors in a very very simple task. Oh, OK, the other explanation is that the group classification schemes used by the two labs are not related.5

I do have a final question on the choice of physiological measures. It appears to be commonplace in the literature for studies to examine how prior presentation of aversive pictures affects the startle response. For example, unlike controls, psychopaths (Justus & Finn, 2007) do not show an enhanced startle to aversive pictures (compared to neutral pictures), while those with borderline personality disorder (Hazlett et al., 2007) show an exaggerated response.

Yet, Oxley et al. looked at these two measures separately. Why? Being more fearful (of certain things) does indeed come with being a xenophobic supporter of socially protective policies, but I would have liked to have learned how the startle response was affected by various pictures, not only of spiders and maggots, but also of undocumented immigrants and gay wedding scenes. But, I think we already know the answer to that question...


1 Oxley et al. (2008) expressly stated that they did not want to use the labels "liberal" or "conservative" -- preferring to describe the more easily startled group as those who
advocate policies that protect the existing social structure from both external (outgroup) and internal (norm-violator) threats.
2 The SciAm article does ask the question:
But what comes first: biology or politics? "It could be working either way," says study author John Alford...
3 See that post for a full list of references, which were removed here for brevity.

4 Or more precisely:
Although the difference was not significant in the bivariate analysis, when the sociodemographic controls were added to better specify the model, the coefficient for blink amplitude was again in the predicted (positive) direction, sizable (standardized regression coefficient = 0.286), and statistically significant (P = 0.03).
5 Amodio et al. simply asked subjects to rate their political attitudes on a -5 (extremely liberal) to +5 (extremely conservative) scale.


Amodio DM, Jost JT, Master SL, Yee CM. (2007). Neurocognitive correlates of liberalism and conservatism. Nature Neurosci. 10:1246-7.

Hajcak G, Foti D. (2008). Errors are aversive: defensive motivation and the error-related negativity. Psychol Sci. 19:103-8.

Hajcak G, McDonald N, Simons RF. (2004). Error-related psychophysiology and negative affect. Brain Cogn. 56:189-97.

Hazlett EA, Speiser LJ, Goodman M, Roy M, Carrizal M, Wynn JK, Williams WC, Romero M, Minzenberg MJ, Siever LJ, New AS. (2007). Exaggerated affect-modulated startle during unpleasant stimuli in borderline personality disorder. Biol Psychiatry 62:250-5.

Justus AN, Finn PR. (2007). Startle modulation in non-incarcerated men and women with psychopathic traits. Pers Individ Dif. 43:2057-2071.

D. R. Oxley, K. B. Smith, J. R. Alford, M. V. Hibbing, J. L. Miller, M. Scalora, P. K. Hatemi, J. R. Hibbing (2008). Political Attitudes Vary with Physiological Traits. Science, 321 (5896), 1667-1670 DOI: 10.1126/science.1157627

Subscribe to Post Comments [Atom]


At September 25, 2008 11:19 PM, Anonymous Anonymous said...

I think your blog is terrific, and I would like to feature you on Wellsphere ( Would you drop me an email?
Good health!
Geoffrey W. Rutledge, MD, PhD

At September 25, 2008 11:47 PM, Blogger The Neurocritic said...

Thanks for the compliment, Dr. Geoff. You must have read one of my other terrific posts, What's With All This Scamming for Free Content?

At September 28, 2008 11:53 PM, Anonymous Anonymous said...

The classic experiment using a startle response was done by clapping hands loudly near a frog; The frog jumped.

The investigator cut off one of the frog's legs and clapped again. The frog jumped.

The investigator cut off the second of the frog's legs and clapped again. The frog jumped.

The investigator cut off the third of the frog's legs and clapped again. The frog jumped.

The investigator cut off the fourth of the frog's legs and clapped again. The frog did not move. From this, the experimenter concluded that the frog was now deaf.


Post a Comment

<< Home

eXTReMe Tracker