Abstract
As societal discussion on the public opinion of science and technology ignites over and over again, understanding where such opinions are rooted is increasingly relevant. A handful of prior studies have suggested personality traits as a root of science and technology attitudes. However, these report mixed findings, and employ small student or convenience samples. This leaves considerable uncertainty regarding personality traits’ relation to attitudes toward science and technology. If in fact stable psychological predispositions play a role, this has considerable implications for science policy and science communication. This article investigates the relationship between the big five personality traits and science attitudes in Germany and the Netherlands. Findings indicate that personality traits are related to science attitudes but only very weakly so, among them openness to experience and negative emotionality are most notably related to science attitudes, whereas extraversion, in contrast to prior studies, shows no relation to science and technology attitudes.
1. Introduction
Scholars of public opinion of science have produced extensive knowledge of individual level predictors of science and technology attitudes. Early literature focused almost singlehandedly on the role of science literacy (Allum et al., 2008; Miller, 1983; Miller et al., 1997; Sturgis and Allum, 2004) as the basis for public attitudes toward science and technology. As this explanation has fallen out of favor due to conflicting findings regarding the efficacy of knowledge provisioning (e.g. Sturgis et al., 2010) and normative issues with the policy implications of the model Wynne (1992), focus has changed to understanding how identity and world views shape public opinion of science. Most prominently such studies show how religious views matter for how people relate to science (Ecklund et al., 2017; Gauchat, 2008; Jelen and Lockett, 2014; McPhetres and Zuckerman, 2018; Simpson and Rios, 2019), and how ideology and partisanship play a role in which groups are more prone to exhibit certain attitudes toward science (Gauchat, 2012; McCright et al., 2013; Nisbet et al., 2015; Pechar et al., 2018). However, while these themes have taken center stage, other individual level characteristics have received less attention. While personality traits are recognized as relevant predictors within neighboring fields of attitudinal research (notably within the study of ideology; Duckitt and Sibley, 2010; Jost et al., 2008), such psychological predispositions have received little attention in the study of public opinion of science and technology.
A handful of prior studies investigate how personality is connected to science attitudes; however, all but one employ either student or convenience samples. Moreover, the central question in previous research has been how personality might affect interest in science and academic success. Findings from student populations might not reflect the general population, and while understanding who is interested in doing science clearly has its merits, general dispositions toward science and technology have much more far-reaching implications for science–society relations. This might be especially problematic, as student samples or samples of individuals who are especially motivated to participate in research might have a profoundly different conceptualization of what science and technology is, making the object of evaluation inconsistent with the conceptualizations of science and technology in the broader population. If some people are more likely to dislike science and technology due to psychological traits, this has deep implications for science communication, institutional outreach, and engagement with science.
Issues of reaching certain sections of the public are well documented in science communication (Dawson, 2018; Humm and Schrögel, 2020). If psychological traits relate to a tendency to have a specific disposition toward science and technology, this might structure reception of science communication and outreach. If, as an example, openness to experience and conscientiousness is related to attitudes toward science and technology, this would mean that emphasizing creativity and innovation of science would be innately more appealing to some, unbeknownst to the communicator. Moreover, if psychological predispositions structure the way in which the public relates to science and technology, attempt to steer public opinion in the direction of science optimism and technology acceptance might prove even harder than has previously been thought, compounding issues of systematic (lack of) uptake in the population. If part of how the public relates to science and technology is grounded in the personality of the individual this could be yet another barrier.
The big five personality traits
Personality traits are “stable (behavioral) tendencies of a person linked to motivational, emotional and cognitive domains over time and across situations” (Montag and Elhai, 2019: 128). Based on consistent findings of a common structure of personality traits throughout the twentieth century, scholars converged on a five-factor model of personality. The “big five” traits included in the model are extroversion, agreeableness, conscientiousness, openness to experience, and negative emotionality (neuroticism; Digman, 1990; McCrae and John, 1992; Montag and Elhai, 2019). While different measurements and conceptualizations exist (e.g. including a sixth honesty–humility dimension; Lee and Ashton, 2018), the dominance of these five traits in the conceptualization of personality stands to this day (Corr and Matthews, 2020). Following Soto and John (2017), individuals high in extraversion are sociable and assertive, those high in agreeableness are trusting and compassionate, high conscientiousness relates to being organized and responsible, negative emotionality entails anxiety and emotional volatility, whereas openness to experience codes for curiosity and creative imagination (see Table 1 in Soto and John, 2017 for a longer list of facets within each trait across operationalizations/conceptualizations of the big five personality traits).
The most cited quality of personality traits is the relatively high stability that they exhibit, both in terms of stability over time, but also in being an access point to the psychological “nature” of individuals by measuring genetically inheritable components of human psychology. The possible genetic roots of personality have received extensive interest, primarily studied in twin studies producing support for claims that personality has a genetic component (see meta studies; Polderman et al., 2015; Sanchez-Roige et al., 2018). With regard to consistency over time, results indicate consistency even across a lifetime (Damian et al., 2019). However, personality does change as individuals mature and experience specific environments (Edmonds et al., 2008). Moreover, the degree to which personality changes in adults varies across individuals, and also seems to have genetic components (Bleidorn et al., 2009). Summing up, personality traits exhibit stability over time and have genetic components, but people do change as they interact with each other and the world.
Does personality predict science and technology attitudes?
Science and technology attitudes refer broadly to attitudes toward science and technology, interest in science and technology, and behavioral intensions regarding science and technology. A broad range of attitudinal constructs have been proposed and investigated through the history of the Public Understanding of Science literature, from the classic measure of promise of science and reservations about science (Miller et al., 1997), to recent developments in trustworthiness of scientists (Besley et al., 2021; Hendriks et al., 2015; Reif and Guenther, 2021) and science-related populism (Mede et al., 2022). Existing knowledge on how personality traits affect science attitudes is, however, scant. Furthermore, study populations are overwhelmingly student or social media samples (see Supplemental Table 1 for an overview).
In the first study to investigate personality traits and science attitudes, Feist (2012) finds that extraversion and openness to experience relate to interest in science, and that all personality traits but negative emotionality relate to some form of science attitudes, across a battery of science related attitudes (science as limited, critical thinking, and required public understanding). Similarly, Tommasi et al. (2021) investigate relationships between the big five and science attitudes (mostly dimensions of interest in science), paranormal beliefs, and competence in science (measures of field-specific science knowledge). Further studies find that personality does not predict views on alternative medicine (Furnham, 2007), two studies indicate that personality predicts belief in “pseudoscience” such as astrology (Andersson et al., 2022; García-Arch et al., 2022), and a recent study finds that COVID-related behavior is predicted by personality traits (Hartmann and Müller, 2023). The specific findings of these studies can be seen in the Supplemental Materials (Table 1). The central finding is that openness to experience is related to more positive stances toward science (technology is less prevalent in these studies), whereas extraversion is related to less positive attitudes toward science. Multiple traits are, however, related less systematically to attitudes toward science. As examples, Feist (2012) finds that conscientiousness relates to higher interest in science and higher propensity to think that science should be understood by the public, and Hartmann and Müller (2023) find that agreeableness is positively related to adherence to COVID measures.
The rationale for why personality traits should be related to science attitudes is less clear in prior studies. For Feist (2012), the main rationale (and focus) is that personality traits relate to higher propensity to be interested in (and pursue) science, leading to higher propensity to also like science. This rationale is broadly present across prior studies, and somewhat mirror deficit model reasoning, in that more interest and knowledge in science should be conducive to liking science, that is, “to know science is to love it” (Allum et al., 2008). A more convincing argument for why attitudes toward science and technology should relate to personality traits is perhaps found in literature on how personality relates to other attitudinal constructs. Most illustrative is the model of Duckitt and Sibley (2010) which argues that personality traits are deep causal drivers of specific worldviews (e.g. low openness relates to seeing the world as dangerous), which drive specific attitudes (e.g. right wing authoritarianism), which lead to the expression of specific ideological positions. Similarly, I expect that personality traits should relate to specific view(s) of the world, and as such the role of science and technology in it. This worldview-mediated model is not incompatible with a model that proposes that personality traits relate to attitudes toward science and technology through a propensity to interact with science. The two may however be seen as representations of the two historically dominant theories of public understanding of science, which would conceivably mediate the relationship between personality and attitudes toward science and technology.
A re-evaluation of the relationship
It seems that personality traits relate to science attitudes, but existing studies investigate this almost exclusively in student and convenience samples, findings are not uniform, samples are generally small, and studies diverge from standard attitudinal measures used in the public understanding of science literature. In this study, I investigate the relationship between the big five personality traits and the most broadly investigated promise of science questions (Miller et al., 1997) in a sample of German and Dutch participants.
To test the expectation that personality relates to attitudes toward science and technology I perform a series of analyses on the World Values Survey Round 6 (Inglehart et al., 2018). This survey round includes both a 10-item battery on the big five personality traits, as well as a variation of the standard promise of science battery. The personality trait measure in this specific dataset has received considerable critique due to poor within-trait correlations outside of western populations (Ludeke and Larsen, 2017), as is often the case outside of WEIRD (Western, Educated, Industrialized, Rich, and Democratic) populations (Laajaj et al., 2019). However, Lu and Cui (2022) recently reexamined the data, and found that issues are largely due to translations errors (notably into Arabic), acquiescence bias and measurement equivalence (the latter isolated to countries not included here). Based on this, I present results for Germany and the Netherlands, the only two western countries in which the big five battery was fielded, to avoid potential translation issues, and I adjust for acquiescence bias.
2. Methods
Participants
This study employs the World Values Survey Round 6 data (Inglehart et al., 2018), using nationally representative samples from Germany (n = 2046) and the Netherlands (n = 1902, total n = 3948). In Germany, data were collected from July to November 2013 in a sample of randomly selected participants through municipality stratified civil register sampling. Data were collected in computer-assisted personal interviews. In the Netherlands, data were collected in December 2013 among members of the LISS (Longitudinal Internet Studies for the Social Sciences) panel, administered by Tilburg University. These two samples are chosen due to data availability, and the data issues elsewhere, described earlier.
Procedure
The relationship between the big five and attitudes toward science and technology is investigated in multivariate ordinary least squares (OLS) regressions including all traits simultaneously and controlling for country, education, age, and gender. The main analysis, indicated by filled dots in Figure 2, predicts science and technology attitudes as an additive measure, but I also investigate the relationship to each of the constitutive variables (hollow dots). See all regression outputs in table form in Supplemental Material. Scripts for reproducing the analyses are available on GitHub.
Materials
10-item BFI
The World Values Survey Round 6 measures personality with the BFI-10 instrument (Rammstedt et al., 2010). Having a short instrument could produce measurement issues, though these are widely used and generally tend to have acceptable psychometric characteristics (Ziegler et al., 2014). As proposed by Lu and Cui (2022), I follow the method from Smith (2004) to correct for acquiescence, by subtracting the mean score across all items within individuals, and then standardizing the individual item-scores within country. Internal consistency is generally acceptable, though agreeableness has a lower than desirable level (α levels: Extraversion = .64, Agreeableness = .42, Conscientiousness = .59, Negative emotionality = .69, Openness to experience = .59).
Science and technology attitudes
To measure science and technology attitudes, I use a group of widely used and long-standing measures of individual views on the promise of science and reservations about science (Miller et al., 1997). These ask respondents to evaluate a series of statements, scaled from 1 to 10. For ease of reading, these are recoded so positive evaluation of science and technology is always high (reversed items are indicated by “rev” in visualizations). Full wording of all six questions are available in Supplemental Table 2. The main results employ a simple additive measure of all of these variables. I rerun analysis with latent variable approaches in Supplemental Table 5, producing virtually identical results.
Control variables
Country is measured with a dichotomous variable (Germany is reference), so is gender (female is reference). Age is measured in years. Educational level is measured with an ordinal variable with nine levels included as-if-continuous. Furthermore, as world-view variables are known to relate to both personality traits (as a consequence) and attitudes toward science and technology (as a predictor), I rerun analysis including these in the models, to evaluate whether these variables crowd out the relationship (Supplemental Table 6, differences presented in Table 7).
3. Results
Before proceeding to the main analyses, I investigate how the six science attitudes relate to the big five personality traits. This can be seen in Figure 1, which shows bivariate relationships between the big five personality traits and the six science attitudes. The relationships are visualized both using LOESS (locally estimated scatterplot smoothing, that is, a smoothed estimation of the bivariate relationship, to see potential non-linear relationships) as well as linear estimation, as is the generally assumed form of the relationship. Moreover, the first row shows non-standardized science attitudes, for readability, while the two last rows show standardized science attitudes, to better display general patterns. As an alternative illustration, correlations can been seen in the correlation matrix in Supplemental Table 3.

Bivariate relationships between the big five personality traits, and the six science attitudes.
The first finding of interest is that the correlations are generally weak. It does not seem that personality traits are a strong predictor of science attitudes, though some traits do have clear and uniform relationships with attitudes toward science and technology. Second, while openness and negative emotionality show consistent (and the strongest) correlations across questions, extraversion, agreeableness and conscientiousness are weakly and/or non-uniformly connected to the individual science attitudes. Moreover, while it is for the most part reasonable to argue for a linear correlation between personality traits and science attitudes, it seems that some personality traits, to the degree that they are related to science attitudes, are so in a slightly non-linear fashion.
Following this, I present OLS regression results. Figure 2 shows estimates and 95% confidence intervals for seven models per personality trait. First for the main analysis of the additive measure, and one model for each individual attitude. These models control for country, age, gender, and educational level and include all five personality traits simultaneously. All variables are standardized (mean = 0, standard deviation = 1). Full regressions are reported in Supplemental Table 4.

Personality trait estimates for OLS regressions.
Agreeableness is only very weakly connected to attitudes toward science and technology. For the additive measure, the regression coefficient is .04. For the items on making lives better, creating opportunities, and making the world better, similar weak correlations are observed, while agreeableness is not related to the questions on depending on science rather than religion, the question about breaking down morals, or perceived importance of science knowledge.
Similarly, conscientiousness is weakly related to science and technology attitudes; the additive measure model returns a coefficient of .05. However, whereas conscientiousness is also not related to perceptions on depending on science rather than faith and breaking down morals, it is most strongly related to perceived importance of interest in science. For the last three items, on how science makes lives, the world, and opportunities better, conscientiousness exhibits weak correlations.
Extraversion does not seem to relate meaningfully to any science attitudes. Only the model on the question on relying too much on science rather than religion, returns a marginally statistically significant result (p < .1), and again only very weakly so.
Negative emotionality has a more substantial correlation with science and technology attitudes. The additive measure model returns a coefficient of –.09. Although still a very weak correlation, this is much more substantial in comparison to the three prior traits. Interestingly, negative emotionality is not related to the question on perceived importance of science knowledge. The strongest correlations are for the perceived consequences of science, making lives better, creating opportunities for future generations and making the world better off.
Finally, openness to experience has the strongest relation to science attitudes. The estimate in the additive measure model is .10. In contrast to negative emotionality, this is most strongly the case for perceived importance of knowledge of science. For the rest of the models openness exhibits smaller coefficients, though once again strongest for the items on how science affects the world, making lives better, creating opportunities and making the world better, rather than the items on the moral consequences of science.
Overall, personality relates only very weakly to this group of attitudes. While neuroticism and openness seem clearly but weakly related to general perceptions of science and technology, this is not the case for the other three personality traits. However, in a series of additional analyses, I find that the relationship is not crowded out by controlling for political and religious views (Supplemental Tables 6 and 7), indicating that the little relationship that seems to exist is not due to value-mediated influence of personality traits. Furthermore, I investigate the proposition that prior investigations among student samples might have misestimated the relationship, due to specific characteristics of such samples. Doing so, in Supplemental Figure S1, by splitting regressions between individuals with/without any university education, reveals some support for such a proposition. While most estimates are quite similar, there are marked differences in the coefficients for negative emotionality, where estimates are considerable stronger for the section of the sample with no university education. This might indicate that general propensity for anxiety produces (slightly) lower evaluations of science and technology, but that this is not the case for individuals who have had significant interaction with science. Interestingly, however, the opposite is the case for agreeableness, which codes for (among other things) a general trusting disposition. This could be of special interest as trust is proposed as one of the main foundations of public perceptions of science and technology in contemporary science communication research.
4. Discussion and conclusion
This study follows a small group of previous studies in hypothesizing that science attitudes should be affected by the big five personality traits. However, whereas previous studies have investigated this almost exclusively among student and social media samples, this study investigates this same connection among the general population and in a much larger representative sample. Two central findings should be highlighted.
First, as in previous studies, I find indications that personality traits have inconsistent relations to different aspects of attitudes toward science and technology. Taken together with prior studies, this might indicate cross-contextual (i.e. different countries or cultures) as well as cross-domain (i.e. different aspects of science and technology) variations in how personality traits matter for public attitudes toward science. Such a finding should be read as a relief to readers concerned that psychological predispositions consistently form attitudes toward science and technology, as this seems to not be the case.
Second, the personality traits most strongly related to science attitudes are negative emotionality and openness to experience. The fact that openness to experience is the strongest predictor of science attitudes is consistent with previous studies. However, the correlations for negative emotionality are so far unique to this study. This is potentially an artifact of prior studies investigating student attitudes, where unease regarding science might conceivably be lower. This proposition is supported by the additional analysis in Supplemental Materials Figure S1, which shows a marked difference between individuals with at least some university education and individuals with no university education. It is puzzling that extraversion shows no clear relation to science attitudes in this study, as introversion is usually associated with scientists (Feist and Gorman, 1998), and extraversion has been demonstrated to be negatively related to interest in science in student samples (Feist, 2012). However, in the general population extraversion does not seem to play an important role, and this trait does not show notable differences between university/non-university respondents. A tentative proposition might be that this inconsistency is the result of a prior focus on personal interest in science. Moreover, consistent findings that conscientiousness matters for academic performance might motivate predictions of conscientiousness playing a central role for science attitudes, as predicted by a science literacy approach. However, this does not seem to be the case, except for perception of importance of interest in science.
While this study represents a considerable improvement on prior similar studies, most centrally by introducing a general population nationally representative sample, it does have considerable limitations. First, the agreeableness measure has a low level of internal consistency, making these specific results less reliable. This is not a trait that has previously been highlighted; however due to its relationship to a propensity to trust, which is a central concept in contemporary science communication literature, this might be of further interest for studies into what leads some (and not others) to trust science, and this shortcoming might hinder discussion based on this specific analysis. Moreover, while this study introduces well-versed attitudinal measures, it might be argued that these do not contain the full breadth of contemporary discussions (e.g. the too much science too little faith question has been under scrutiny; Roos, 2014), a replication with a more comprehensive and updated list of attitudinal components might be of relevance.
While any single study cannot definitively settle if personality traits relate to public attitudes toward science and technology, this study provides a first look at such correlations in a representative general population sample. Centrally, findings indicate that personality traits are clearly not the rock on which public opinion of science and technology is built. Although some personality traits do relate to some science and technology attitudes, they do so weakly. This might be expected for distant predictors, but still this tells a tale of caution when arguing that psychological predispositions are core features to the formation of science attitudes. If some people simply do not like science, it seems that this is only weakly rooted in personality traits. Personality traits and other (relatively) stable psychological factors might play a role in public perceptions of science, but based on the above findings only to a minor and, crucially, non-uniform degree.
Supplemental Material
sj-docx-1-pus-10.1177_09636625231220341 – Supplemental material for What if some people just do not like science? How personality traits relate to attitudes toward science and technology
Supplemental material, sj-docx-1-pus-10.1177_09636625231220341 for What if some people just do not like science? How personality traits relate to attitudes toward science and technology by Simon Fuglsang in Public Understanding of Science
Footnotes
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.
Author biography
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
