This past week the conservative blogosphere was alive with finger pointing and pseudo analysis of a research paper published in the December 2011 edition of the American Journal of Political Science. The paper was titled, Correlation Not Causation: The Relationship between Personality Traits and Political Ideologies. The authors of this paper were Brad Verhulst, Post-Doctoral Fellow, Lindon J. Eaves, Distingushed Professor, and Peter K. Hatemi, Associate Professor Political Science and Microbiology.
The reason for all the uproar over some obscure academic article is additional information has come to light that shows the authors had misunderstood their data and the conclusions they reached were exactly 180 degrees off from what the data showed. Since its original findings seemed to reflect poorly on the conservatives they are now beating the drum on the new conclusions.
So let’s talk about this a minute.
The study set out to show that personality traits do not cause a person’s choice in political ideology, as previous studies had assumed, but political attitudes develop much earlier in life and there is a direct correlation between the two that can be traced back to the individual’s genetic factors.
To help show this they took data from a fairly large scale study of twins known as the Mid-Atlantic Twins Registry (MATR) and an AARP mailer. To assess the personality traits, they used the Eysenck Personality Questionnaire (EPQ). Let’s pause here to discuss Mr. Eysenck.
Hans Eysenck was born in Germany in 1916, but fled to England in the 30’s to escape Nazi persecution. He received his PhD in Psychology from the University of London in 1940. He is widely cited for his work in linking personality to the genetic make-up of the the individual.
Eysenck developed a definition of personality that suggests it is the “sum of the actual or potential behavior patterns of the organism, as determined by heredity and environment it originated and develops through the functional interaction of the four main sectors the behaviors-patterns are organized.”[i] He goes on to classify personality into three dimensions: Psychoticism, Extraversion, and Neuroticism, or PEN for short.
Extraversion vs Introversion – Extroverts are toughminded, impulsive, have a tendency to be outgoing, desire for novelty, prefer vocations involving contact with people, have a tolerance for pain, and their performance is enhanced by excitement. Introverts, on the other hand are tenderminded, introspective, serious, easily aroused but restrained, inhibited, with preference for solitary vocations, and their performance is degraded by excitement.
Psychoticism vs Super Ego – This was the last dimension developed by Professor Eysenck and is, therefore, the least defined, but in summary at the two ends of the spectrum you have either a non-conformist who is tough minded and unconcerned with the welfare of others, or a warm sensitive individual who makes concern for others a priority.
Finally, we have Neuroticism. Eysenck classified neurotics as unstable individuals governed by irrational fears and concerns. The opposite end of the spectrum is therefore a stable individual who is calm and even-tempered.
Figure 1 summarizes the three spectrum's that Eysenck believed make up human personality.-->
|Figure 1: Eysenck's PEN Model|
The “Correlation, Not Causation” study used the PEN model to help explain the linkage to personality the separates the liberal from the conservative. They also used something called the Five Factor Model or FFM. According to the researchers this is another popular modeling tool that has been used to show a correlation between personality aspects and political positions and they talk about a complex relationship between the FFM and EPQ regarding a characteristic knows as “Openness to Experience,” which other studies have suggested is a strength of those who are politically liberal.
In the study the authors conducted four sets of analysis.
The first was to demonstrate a significant relationship between personality traits and political ideology. In the authors words, this study demonstrated “there are several substantively significant relationships between the personality traits and political ideology dimensions. Most notably, P [psychoticism] is substantially correlated with conservative military and social attitudes, while Social Desirability is related to liberal social attitudes, and Neuroticism is related to liberal economic attitudes.” If the data they used to determine this was assessed 180 degrees out then it is safe to assume that social desirability is a conservative value while P actually correlated with liberal social attitudes, i.e. risk taking, a lack of concern for others, and tough-mindedness, leaving the conservatives as the ones who are most concerned with the welfare of others. With regard to Neuroticism, it would be most correlated with conservative economic attitudes. The authors were puzzled by this relationship, and the failure to correctly evaluate their data would explain that confusion.
In the other analysis they attempt to confirm previous studies, in the case of the third and fourth analysis examine the relationship between personality and political attitudes. I am not sure how the use of the core data affects the understanding of causal relationships, and it seems relatively unimportant to the overall assessment of the current social media frenzy.
So we come to the question of the study. Do we see in practical application anything that suggests that the data, once correctly interrupted is reflected in the real political world?
Are Liberals more likely to be: tough-minded, non-conforming, risk takers with little concern for the rights of others, and occasionally exhibiting anti-social behavior but overall they feel good about themselves?
Are Conservatives more likely to exhibit unsubstantiated fears, obsessive behaviors, fear and anxiety over fiscal policies, but are at the same time sensitive, warm and concerned about the rights of others?
Beats me… I only have the illusion of knowledge… I don’t claim to know everything -- unlike most of the people on the internet.