A self-assessment Internet survey was conducted by Brack and Zhang on Neuro-politics.com. The survey collected 490 questionnaires, which were subsequently cleansed for 48 cases of probable duplicate or incomplete responses. This left 442 usable questionnaires, of which 218 indicated support for George W. Bush, 195 for John Kerry, and 29 for unspecified other candidates. This survey period was January 28 through March 4, 2005. The survey questions were adapted from similar left-right brain surveys, with slight variations. Additional questions were included to determine family birth order, political affiliations, and strength of religious belief.

The survey indicated a statistically significant left hemispheric bias, (p < .02), for subjects that described themselves as Conservative as compared to both Moderates and Liberals. (For this statistic, we have combined the Very Conservative into the Conservative cohort. We have also combined the Very Liberal into the Liberal cohort). When not combined, this left hemispheric bias was exacerbated in subjects that described themselves as Very Conservative.

There was a statistically significant right hemispheric bias, (p < .01) for subjects that described themselves as Liberal, as compared to Conservatives and Moderates. (For this statistic, we have combined the Very Conservative into the Conservative cohort. We have also combined the Very Liberal into the Liberal Cohort). When not combined, this right hemispheric bias was exacerbated in subjects that described themselves as Very Liberal.

Liberals had a slightly lower left hemisphere score than Moderates, but this was not statistically significant. Moderates exhibited a slightly higher right hemisphere score than Conservatives, but this too was not statistically significant.

Moderates exhibited the smallest mean difference between left and right responses, and also had the lowest average left + right responses.

The survey results are consistent with the hypothesis of general left hemispheric biases for Conservatives and general right hemispheric biases for Liberals, and more neutral biases for Moderates.

Relative to Conservatives, Liberals tend to exhibit a more right-brained orientation for language processing and simple arithmetic processing, reality model building, creativity and artistic expression, and emotional processing.

Relative to Liberals, the left-brain oriented Conservatives tend to exhibit enhanced literal language processing skills, exact arithmetic calculation skills, a greater tendency to build binary reality models of the world (i.e., "Good" and "Evil"), a greater tendency for belief-bias, and an enhanced time-vector orientation.

Moderates exhibit a more blended attribute mix, with equivalent simple arithmetic skills as the Conservatives, but a lesser ability in literal language processing that is more comparable to Liberals. Moderates show enhanced time-vector orientation when compared to Liberals, and a lower tendency to build binary reality-models of the world than Conservatives. Moderates tend to exhibit less belief-bias than the Conservatives. Moderates generally fall in between Conservatives and Liberals in many of the questions.

Those indicating support for George W. Bush had a statistically significant left hemispheric bias, (p < .02). Those indicating support for John Kerry had a statistically significant right hemispheric bias (p < .01).

The survey indicated a statistically significant left hemispheric bias, (p < .01), for subjects that described themselves as Very Religious as compared to both Not Religious and A Little Religious.

There was also a statistically significant right hemispheric bias, (p < .02), for subjects that described themselves as Not Religious as compared to Very Religious

There was also a statistically significant right hemispheric bias, (p < .06), for subjects that described themselves as a Little Religious as compared to Very Religious.

The Very Religious mirror the Conservative cohorts quite closely with general left hemispheric biases. They exhibit a similar tendency for belief-bias and binary reality models, exact simple arithmetic skills, enhanced time-vector orientation, enhanced literal language processing skills, and a lesser tendency for right-brained emotional processing.

The Not Religious mirror the Liberals quite closely with their general right hemispheric biases. They exhibit similar deficits at literal language processing and exact simple arithmetic processing when compared to the Very Religious. They exhibit a slightly higher tendency than the Liberals for belief-bias and binary reality model building, but much less than the Very Religious. They exhibit less right brain emotional processing attributes than the Liberals, but still higher than the Very Religious. Breaking pattern, they are less likely to exhibit affective language.

The Little Religious have an interesting characteristic, which is a higher tendency towards right hemisphere emotional processing than either the Very Religious or the Not Religious. They have a slightly more difficult time than the Very Religious at literal language processing, and similar deficits in simple arithmetic as compared to the Not Religious. They fall in between the Very Religious and Not Religious in belief-bias and binary reality-model building.

However, Religious and Political cohorts need to cross-referenced, as they are not adequately analyzed separately. We will do that in a later installment of our study.

The most notable statistics of this survey were derived from taking the mean left brain score and subtracting the mean right brain score for each political affiliation:

This remarkable result does not necessarily imply that Liberals are right-brain dominant and Conservatives are left-brain dominant. This survey never intended to resolve left versus right brain dominance, nor even resolve what dominance means. This survey was focused on a set of mental attributes that have demonstrated hemispheric asymmetries.

1 A. Left-Brain Scores By Political Affiliation

Graphically:

There was a notable increase in mean left responses as one moved from left to right in the political spectrum. Very Liberal, Liberal, and Moderates all showed no statistically significant differences compared with each other. Conservatives showed a slightly significant (p < .26) variation with Moderates, and wider variations with Liberals (p < .09) and Very Liberals (p < .07). Very Conservatives showed the widest variation with all cohorts, even when compared to Conservatives (p < .04). Compared to Liberals, Very Liberals, and Moderates---Very Conservatives exhibited a large Left-bias (p < .01). See Table 3.

1 B. Right-Brain Scores By Political Affiliation

The right-hemisphere pattern looks very much like the opposite of the left-hemisphere pattern of Figure 2. There was a significant elevation in the mean right-hemisphere responses in the Very Liberal and Liberal cohorts when compared to Moderates, Conservatives, and Very Conservatives (all p < .01). The Very Liberals had the highest mean responses, with a small but possibly statistically significant (p < .35) that they even diverge from the regular Liberals. Conservatives, Moderates, and Very Conservatives have very similar response patterns when compared to each other. See Table 5.

Taking the Left-Hemisphere score and subtracting the Right-hemisphere score for each respondent, we compute a Difference variable that is analogous to a Hemispheric Bias for the mental attributes we surveyed. This is not to be confused with brain dominance, which was not explicitly measured in this survey. Our Hemispheric Bias is negative for right brain oriented respondents, and positive for left brain oriented respondents. Very Liberals exhibited a significant rightward bias. Regular Liberals exhibited a rightward bias, and moderates had a slight leftward bias. Conservatives exhibited a slightly elevated leftward bias compared to Moderates, and Very Conservatives were strongly left biased. Table 6 mean responses are graphed in Figure 1.

Table 7 shows the comparisons of the political cohorts along with their Z-values and corresponding probabilities that the differences are due to random sampling. All cohorts show differences from each other that are at least mildly statistically significant. Liberals and Very Liberals are the closest two cohorts (p < .39), followed by Conservatives and Moderates (p < .36). Conservatives and Very Conservatives (p < .15), were further apart than Conservatives and Moderates. Very Liberals and Very Conservatives show the widest difference, with a large Z-value of 5.34

2 A. Left and Right Brain Scores by Presidential Preference

The following tables do not include the missing and "other" response statistics in the analysis. Bush and Kerry supporters showed a hemispheric bias consistent with the left and right hemispheric orientations by Political Affiliation cohorts. Note the higher variation on the right scale.

For the sake of brevity, we will not chart hemispheric comparisons, but go directly to difference analysis between Bush and Kerry supporters. Bush supporters show a positive or left-hemispheric bias, while Kerry supporters exhibited a corresponding right-hemispheric bias.

Graphically:

3 A. Left and Right Brain Scores by Religious Preference

Religious preference also shows a significant asymmetry in left hemisphere scores. Very Religious respondents had a similar mean left score as our aggregated Conservative + Very Conservative cohort. A Little Religious respondents had a similar left brain score as our Moderates, and the Not Religious had a mean score similar to our Liberal + Very Liberal cohort.

Graphically:

The right hemisphere responses show an inverse trend as compared to the left hemisphere scores. Very Religious people drop more than one full point in their right hemisphere scores (5.57), as compared to their left hemisphere scores (6.8). A Little Religious maintain their middle position in right hemisphere scores (like they did in the left hemisphere scores), and Not Religious people moved from the bottom of the left scale to the top of the right. See Table 12.

Graphically:

Left brain scores show a statistically significant (p < .01) that the Very Religious have different population means from Little Religious and Not Religious. Interesting to note that the Z-value is greatest when we compare VR to NR. There is a slightly statistically significant difference between Little Religious and Not Religious (p < .37).

Right brain scores also show statistically significant differences between VR and LR (p < .06) and NR (p < .02). No statistical significant difference is noted between LR and NR (p < .60).

If we subtract the mean right score from the mean left score, and take this value and compare our religious cohorts, we get Table 15.

We again see a statistically significant variation (p < .01) that the Very Religious have different population means than either the LR or NR cohorts. The Very Religious have Left minus Right difference scores similar to the combined Conservative + Very Conservatives. The Little Religious fall in the middle, with a slightly statistically significant variation in difference scores with the NRs (p < .31). Figure 7 shows the large left score bias in the Very Religious.

We have focused primarily on statistical analysis of summary data, but this doesn't tell us much about how our political cohorts vary from each other. The survey consisted of 29 left or right questions, with two control questions placed at the end to detect completeness of the survey. We will now focus on the questions that were mostly responsible for separation of the sample means, ones that contributed to a 0.1 or greater difference between Conservatives and Liberals

Question 3: I believe there is a right and wrong way to do everything.

A "Yes" to this was scored as a left-hemisphere response. This is predominately a left-hemisphere binary categorization. Binary views of "Right" and "Wrong" or "Good" and "Evil" are left hemisphere belief-biased events. Belief-biased events have been localized in the left temporal lobe by V Goal and RJ Dolan (1). Interestingly enough, Goal and Dolan also found belief-neutral reasoning originated from a bilateral parietal lobe system, and activation of the right lateral prefrontal cortex was evident when subjects inhibited a belief-biased response. Note the progressive elevation of responses as one goes from both the VL to VC political spectrum and as one goes from NR to VR in the religious spectrum.

A "Yes" to this was scored as a left-hemisphere response. Ornstein, Herron, Johnstone, and Swencionis (2) found that EEG activity in the right hemiphere predominates when processing fiction, EEG in the left hemisphere predominates when processing science textbooks. Our religious spectrum is less clear than our political spectrum, and no cohort in the religious spectrum exceeds 50%.

Question 1: I always wear a watch.

A "Yes" to this was scored as a left-hemisphere response. Watch or clock processing are predominately left hemisphere events. Rotenberg(3) has theorized that the vector of time is a left hemisphere function, as part of its general tendency of building unambiguous and organized reality models.

Question 9: I am pretty good at arithmetic.

A "Yes" to this was scored as a left-hemisphere response. Stanescu-Cosson et al (4) provided evidence that exact calculation has been associated with the left inferior prefrontal cortex and the bilateral angular regions. Approximation involves a bilateral prefrontal system. Note the high score for the Very Religious.

Question 28: When I'm confused, I usually go with my gut instinct.

A "Yes" to this question was scored as a right-hemisphere response. In retrospect, this is questionable. Goel et al(5) found that when a logical argument results in a belief-logic conflict, the nature of the reasoning process is changed by the recruitment of the right prefrontal cortex. Our results may infer that normal left-hemisphere belief-based logical conflicts, presumably self-assessed as "confusion", result in reliance on the intervention of right hemisphere processing, also self-assessed as "instinct". Therefore, the question may not be detecting hemispheric dominance as much as what happens when the left-hemisphere cannot resolve an internal logical inconsistency. This might explain why Conservatives and Moderates had higher "Yes" responses, although the low response percentages seem to impact only the Very Liberal and Not Religious.

Question 25: I lose track of time easily.

A "Yes" to this question was scored as a right-hemisphere response. This is a similar question asked in the opposite to "I always wear a watch". We see a noted elevation in Very Liberal, Liberal and Not Religious cohorts. As we discussed previously, the vector of time is a left-hemisphere event.

Question 4: I find it hard to follow directions exactly.

A "Yes" to this question was scored as a right-hemisphere response. The brain's left-hemisphere bias for processing language syntax is well documented. This question implicates written and/or verbal instructions being more difficult for the Very Liberal, Liberal, and Not Religious cohorts. Right-hemisphere language processing is more effective with short phrases, and has a much more limited short-term memory than the left-hemisphere (7).

Question 16: I am musically inclined.

A "Yes" to this question was scored as a right-hemisphere response. This is questionable, in that hemispheric dominance for melody recognition varies based on level of exposure and has been shown by Bever and Chiarello (6) to be different for the musically trained (left-hemisphere oriented analysis) and untrained (right-hemisphere oriented). Formulation of this question does not sufficiently delineate a potential hemispheric bias, and may explain the mixed results. Interesting to note is the religious spectrum, showing almost no variance in cohorts.

Question 5: I find that sticking to a schedule is boring.

A "Yes" to this was scored as a right-hemisphere response. Recent neurochemical evidence by R Kurup and P Kurup (8) indicates asymmetrical neurotransmitter patterns in creative versus non-creative people. This pattern in non-creative individuals correlated with that obtained in left hemispheric chemical dominance. The pattern in creative individuals correlated with right hemispheric dominance. This analysis presumes that creative people find that schedules are more "boring" than non-creative people. There is a significant downward tendency in the both the political and religious spectrums as one goes from left to right.

Question 8: If I don't know which way to turn, I let my emotions guide me

A "Yes" to this was scored as a right hemisphere response. The right hemisphere generally predominates in emotional processing. Right hemispheric bias in activation in autonomic responses suggests that the physiological response system rather than the perceptual/cognitive system is the locus of the right hemisphere superiority for emotion (9). In no cohort does the response rate exceed 50%. Lows in both the Very Conservative and Very Religious. The Little Religious break the pattern, being the highest cohort in the religious spectrum.

Question 18: When I talk, I gesture a lot with my hands

A "Yes" to this was scored as a right-hemisphere response. Even though motor planning of hand movement predominates in the left-hemisphere, Lausberg, Davis, and Rothenhausler (10) found that gestures with emotional connotation are generated predominately in the right-hemisphere. While the liberal wing of the political spectrum had elevations in hand gestures, the Very Religious and Little Religious scored higher than the Not Religious. This interesting anomaly will be discussed in a later installment of this study.

Question 11: I'm frequently late getting places

A "Yes" to this question was scored as a right-hemisphere response. This is a similar question to 25, "I lose track of time easily". Again the liberal wing of the political spectrum is elevated relative to the Moderates and Conservatives, and the Not Religious and Little Religious are almost double the Very Religious cohort.

Question 24: I've considered becoming a poet, a politician, an architect, or a dancer

A "Yes" to this was scored as a right-hemisphere response. Poetry is more metaphoric and non-literal, both well-documented right hemisphere biased functions. Architecture is visuospatial, another right-hemisphere oriented function. Dancing is indicative of motor coordination and grace of performance of nonverbal behavior as a whole are under the control of the right hemisphere (12). Politician is questionable, and we see a break in the pattern for Very Conservatives, which may indicate a "strength" of political belief effect on selection of this particular occupation.

Question 27: I like to draw

47% of the Very Liberal cohort scored a -2 or less (right hemisphere tendency) in the Left minus Right statistic. This compares to just 16% of the Very Conservative cohort scoring -2 or less. Conversely, 48% of the Very Conservatives scored +2 or more (left hemisphere tendency). This compares to just 19% of the Very Liberals scoring +2 or more. While the difference between the right and left hemispheric scores between the Very Liberals and Very Conservatives is significant, a simple theory of hemispheric bias will not be a very effective predictor of whether a particular individual will be Conservative or Liberal. We still find Very Conservatives with right hemisphere biases, and Very Liberals with left hemisphere biases.

Obviously, we are missing something. Our initial thinking prior to conducting this survey was that the major difference between Liberals and Conservatives was that the Liberals had a greater tendency for "abstract social empathy", which was empathy for people without direct contact. We even speculated that the source of this "abstract social empathy" was in the right orbital prefrontal cortex. But there was no attempt to determine "abstract social empathy" tendencies in this survey.

The inclusion of "abstract social empathy" would not likely be the missing piece of our Liberal-Conservative puzzle anyway. We would still find conservatives with high "abstract social empathy", and liberals with none.

This points heavily to other factors in the make-up of Conservatives, Liberals and Moderates. We can only speculate what those other factors are. There is a significant link between parent and child with regard to political and religious beliefs, which may extend beyond the genetic and into the developmental. There is already long-standing evidence of the role of gender in political decision-making, and this points to the steroid hormones---such as testosterone. The role of androgens on brain development has been well documented and shows evidence of certain lateralizations (13), which may be relevant in the study of hemispheric asymmetries and political tendencies.

An adequate analysis of the data cannot take place without breaking it down further into Political-Religous cohorts. The hemispheric biases of a Very Religious Moderate are indeed different than a Little Religious Moderate, and this holds true across the political spectrum, as we shall see in a later installment of this study. In Part 2, we will focus more on analysis by gender.

5. Goel V, Buchel C, Frith C, Dolan RJ. Dissociation of mechanisms underlying syllogistic reasoning Neuroimage. 2000 Nov;12 (5):504-14.

6. Bever, Thomas G., Robert J. Chiarello. Cerebral Dominance in Musicians and Nonmusicians,Science 185 (1974): 537-539.

7 Zaidel, Eran, Language and the Disconnected Hemisphere, Elsevier's Encyclopedia of Neuroscience, 2nd Edition p1028.

9. Spence S, Shapiro D, Zaidel E. The role of the right hemisphere in the physiological and cognitive components of emotional processing. Department of Psychology, UCLA.

10. Lausberg H, Davis M, Rothenhausler A. Hemispheric specialization in spontaneous gesticulation in a patient with callosal disconnection. Neurology 2002 Sep 24 59(6) 893-897

12. Springer, S.; Deutsch, G. Left brain, right brain. New York: W. H. Freeman and Co.; 1985.