First published February 27, 2021.
A 2021 study led by psychologist Dr Leor Zmigrod of the University of Cambridge found that people who strongly endorsed extreme pro-group actions, including violence in the name of their group, tended to share a recognisable psychological profile: weaker working memory, slower and more cautious perceptual decisions, and personality traits linked to impulsivity and sensation-seeking. The study measured performance on mental tasks and questionnaires, not brain structure, and it showed associations, not causes.
Headlines at the time summarised it as "people with extremist views are less able to do complex mental tasks". That captures part of the finding but misses much of what makes the research interesting. Here is what the scientists did, what they found for different kinds of belief, and what the results can and cannot tell us.
How the study worked
The paper, "The cognitive and perceptual correlates of ideological attitudes: a data-driven approach", was published in Philosophical Transactions of the Royal Society B by Zmigrod with Ian Eisenberg, Patrick Bissett, Trevor Robbins and Russell Poldrack. It built on a large Stanford University project from 2016 and 2017, in which hundreds of American adults completed 37 cognitive tasks and 22 personality surveys.
In 2018, 334 of those participants, aged 22 to 63 and just under half of them women, returned to fill in 16 further surveys about their ideological attitudes and how strongly they held them. The cognitive tasks were deliberately neutral. They involved things like sorting coloured shapes, holding lists of words in mind and making rapid decisions about moving dots on a screen, with nothing political in them at all.
The researchers then used computational models, including drift-diffusion models of decision-making and Bayesian statistics, to look for links between how people thought, what kind of personalities they reported and what they believed.
What they found for each type of belief
One of the study's strengths is that it did not lump all beliefs together. According to the University of Cambridge summary, the profiles differed:
- Political conservatism and nationalism were linked to greater caution in quick decisions (slower but accurate responses), a stronger tendency to discount delayed rewards and slightly reduced strategic information processing. On personality measures, these participants scored higher for goal-directedness and reward sensitivity, and lower for social risk-taking.
- Dogmatism, meaning a fixed worldview resistant to evidence, was linked to slower evidence accumulation in perceptual tasks, along with lower agreeableness, more impulsivity and more willingness to take ethical risks.
- Religiosity showed cognitive patterns similar to conservatism, together with higher agreeableness and a heightened perception of risk.
- Extreme pro-group attitudes, including support for ideologically motivated violence, showed a blend of the conservative and dogmatic signatures: cautious, slower perceptual processing and weaker working memory, combined with impulsivity and a taste for risky, thrilling experiences.
That last group is the one behind the headlines. Importantly, the extremism measure was not tied to one side of politics; it captured willingness to take extreme action for whichever group or cause a person identified with.
How much the psychology explained
The team compared how well demographics, such as age, sex and education, could predict ideology against a model that also included cognitive and personality measures. For conservatism and nationalism, demographics alone explained under 8% of the variation between people; adding the psychological measures raised that to about 32.5%. For dogmatism the figures rose from about 1.5% to 23.6%, and for religiosity from about 2.9% to 23.4%.
In other words, the combined model was several times better than demographics alone, while still leaving most of the variation unexplained. Zmigrod described the aim as understanding why particular individuals are more susceptible
to extreme ideas, which is a question about vulnerability rather than blame.
What the study does not show
Several common readings of this research go beyond the evidence:
It does not show that extremists are unintelligent. The tasks measured specific processes such as working memory and decision speed, not general intelligence, and the differences were averages across groups. Many people with the same cognitive profile hold moderate views.
It does not show cause and effect. The data are correlational. Cognitive style might make some ideologies more appealing, but strongly held beliefs, life stress or other factors could also shape how people perform on tasks.
It does not map the brain. No brain scans were involved. Talk of "extremist brains" in some coverage was shorthand, not a description of the method.
It is not a screening tool. The authors described the work as early-stage. A study of 334 Americans cannot reliably identify a dangerous individual, and using such measures to label people would raise serious ethical problems. The researchers saw possible value in understanding who might need support, not in policing beliefs.
It is also worth being careful with wider claims. Earlier research on intelligence and political views is mixed, and associations, where found, are usually modest and depend heavily on how ideology is measured. The Cambridge findings come from a single US sample and need to be repeated in other countries and populations.
Why this research matters
Much research on radicalisation focuses on social factors: grievances, propaganda, online communities and personal networks. Those remain central. What this study adds is the suggestion that individual differences in how people process information may make some more receptive to ideologies that offer clear, simple answers and demand total loyalty.
Zmigrod's broader work has explored a related idea, cognitive flexibility, the ability to switch strategies when the rules change. People who find that harder may feel more comfortable with rigid belief systems. If that holds up, it points towards practical responses: education that builds tolerance for uncertainty, media literacy, and programmes that help people encounter opposing views without feeling threatened.
Practical takeaways
- Read headlines about "the extremist mind" with care: check whether a study measured the brain, behaviour or survey answers, and whether it found causes or correlations.
- Remember that group averages say little about any one person, including people you disagree with politically.
- Habits that slow down snap judgements, such as seeking out evidence that might prove you wrong, are useful for everyone, whatever their politics.
Sources: University of Cambridge, "Psychological signature for the extremist mind uncovered", 22 February 2021; Zmigrod et al., Philosophical Transactions of the Royal Society B (2021); Zenodo data record for the study; earlier coverage by The Guardian.
COMMENTS