key: cord-0759658-wgsaawu9
authors: Rudloff, Jan Philipp; Hutmacher, Fabian; Appel, Markus
title: Beliefs about the nature of knowledge shape responses to the pandemic: Epistemic beliefs, the Dark Factor of Personality, and COVID‐19–related conspiracy ideation and behavior
date: 2022-03-07
journal: J Pers
DOI: 10.1111/jopy.12706
sha: 82a1ffd0528bc95647848bc994992a2d63929e7c
doc_id: 759658
cord_uid: wgsaawu9

OBJECTIVE: Global challenges such as climate change or the COVID‐19 pandemic have drawn public attention to conspiracy theories and citizens' non‐compliance to science‐based behavioral guidelines. We focus on individuals' worldviews about how one can and should construct reality (epistemic beliefs) to explain the endorsement of conspiracy theories and behavior during the COVID‐19 pandemic and propose the Dark Factor of Personality (D) as an antecedent of post‐truth epistemic beliefs. METHOD AND RESULTS: This model is tested in four pre‐registered studies. In Study 1 (N = 321), we found first evidence for a positive association between D and post‐truth epistemic beliefs (Faith in Intuition for Facts, Need for Evidence, Truth is Political). In Study 2 (N = 453), we tested the model proper by further showing that post‐truth epistemic beliefs predict the endorsement of COVID‐19 conspiracies and disregarding COVID‐19 behavioral guidelines. Study 3 (N = 923) largely replicated these results at a later stage of the pandemic. Finally, in Study 4 (N = 513), we replicated the results in a German sample, corroborating their cross‐cultural validity. Interactions with political orientation were observed. CONCLUSION: Our research highlights that epistemic beliefs need to be taken into account when addressing major challenges to humankind.

It is crucial for democratic societies that their members act upon evidence, not least when faced with global challenges such as climate change or the COVID-19 pandemic (Lewandowsky et al., 2020) . The popularity of conspiracy theories and the widespread failure to follow behavioral guidelines informed by science has fueled academic research investigating the antecedents of such thinking and behavior (e.g., Boot et al., 2021; Pennycook et al., 2020) . Individuals' convictions or worldviews about how one can and should develop a sense of what is true could be a key to understanding post-truth phenomena (Hyman & Jalbert, 2017; Lewandowsky et al., 2017; Scheufele & Krause, 2019) . Our focus here is on individual differences in epistemic beliefs, that is, people's concepts about knowledge and evidence (Schommer, 1990) . Among other aspects, epistemic beliefs refer to the validity of truthiness (Colbert, 2005) -the belief that truth rightfully derives from one's gut feeling rather than facts. Building upon a three-dimensional framework of epistemic beliefs (Garret & Weeks, 2017) , we examined the Dark Factor of Personality as an antecedent to the endorsement of epistemic beliefs and COVID-19 conspiratorial thinking and the (non-)adherence to WHO health behavioral guidelines during the pandemic as consequences of epistemic beliefs. We start with a brief introduction on epistemic beliefs.

Imagine reading the abstract of a recently published study on COVID-19. Furthermore, imagine that you find the authors' results and conclusions intuitively plausible. Do you believe that you can trust your gut feeling-or do you rather find it necessary to take a closer look at the study before drawing any conclusions? Also, do you think that science provides objective facts or do you rather hold that scientific conclusions are influenced by those in power? These questions refer to your epistemic beliefs. Epistemic beliefs can be defined as beliefs about the nature and generation of knowledge (Muis, 2007) . The concept of epistemic beliefs emerged in educational psychology (see, e.g., Hofer & Pintrich, 1997; Kuhn et al., 2000; Schommer, 1990) , but has since also been adapted to study the general public. In contrast to the concept of motivated reasoning (Kruglanski, 1996; Kunda, 1990) , which is the process of producing justifications or decisions based on individual motives, goals, and attitudes instead of evidence, epistemic beliefs refer to one's general attitude toward the concept and generation of knowledge itself (Hornsey et al., 2020) . It has been demonstrated that epistemic beliefs are related to the accuracy of peoples' opinions (Garrett & Weeks, 2017) . What does that mean?

To capture the impact of epistemic beliefs on cognition and behavior, three different aspects need to be distinguished (see Garrett & Weeks, 2017) . First, the degree to which people have Faith in Intuition for Facts, that is, the degree to which people believe that they can trust their gut feeling when evaluating information. Intuition can be an important source of knowledge (see, e.g., Damasio, 2005; Kahneman, 2011) , especially when taken as a starting point for further careful and thorough consideration. If faith in one's intuition is not accompanied by analytic thinking however, people tend to ignore and disregard existing evidence, which can lead to severe misperceptions (e.g., .

Second, the degree to which people have a Need for Evidence, that is, the degree to which people believe that their opinions need to be based on externally validated data. People with a high Need for Evidence will try to ensure that their opinions align with the known facts. On the contrary, people with a low Need for Evidence hold opinions that are driven by their ideological convictions, even if they know that these convictions conflict with the current scientific consensus (e.g., Garrett et al., 2016; Hindman, 2009) .

Third, the degree to which people believe that "facts" are shaped by those in power, that is, the degree to which people believe that Truth is Political. In the social sciences, it has been emphasized that the generation of knowledge is always embedded in historical and societal circumstances (e.g., Hacking, 1999) . Such a social constructionist perspective can easily be misunderstood as entailing a disregard for the truth (Holtz, 2020) . People who believe that facts do not exist independently from the political context tend towards this misunderstanding (Garrett & Weeks, 2017) . Based on the observation that there are different interpretations of reality and that the scientific consensus shifts over time, it is concluded that "truth" is nothing but a matter of power.

To sum up, there is a distinct set of epistemic beliefs that we call post-truth epistemic beliefs because they shield individuals from questioning their opinions and from engaging in a rational discourse. In other words, people with a strong Faith in their Intuition for Facts, a low Need for Evidence and a strong conviction that Truth is Political will show little inclination to commit to "the unforced force of the better argument" (Habermas, 1996, p. 305) . Instead, they will deliberately choose to believe what they want to be true. In the following, we argue that epistemic beliefs are closely connected to and an expression of a broader personality disposition, the Dark Factor of Personality.

The Dark Factor of Personality (D) is defined as "the general tendency to maximize one's individual utilitydisregarding, accepting, or malevolently provoking disutility for others-accompanied by beliefs that serve as justifications" (Moshagen et al., 2018, p. 657) . Utility refers to any form of material success or hedonistic feelings such as power or pleasure. The concept of D can be understood as an equivalent to G, the core factor of intelligence, explaining common variance between dark traits such as egoism, Machiavellianism, and psychopathy (Moshagen et al., 2018) . Individuals high in D embrace a relativist and cynical worldview, which enables them to bend moral values and to refrain from injunctive norms, whenever it suits their agenda (e.g., Ajzen, 1991; Jonason et al., 2015; Moshagen et al., 2018 Moshagen et al., , 2020 Zeigler-Hill et al., 2020) .

Thus, it is to be expected that the stronger D, the stronger is the tendency to approve external information that justifies an individual's antagonistic, malevolent or socially aversive behavior while disregarding information criticizing it. We argue that epistemic beliefs serve as tools to construct and maintain convictions that serve as justifications for this behavior. Furthermore, we assume that individuals high in D use epistemic beliefs to fend off information contradicting their worldviews even if these worldviews are not directly linked to justifying antagonistic, malevolent or socially aversive behavior, as threats to any worldview may cause disutility. In addition, as individuals high in D are likely to hold positive, but fragile self-concepts (e.g., Doerfler et al., 2021) , epistemic beliefs serve as a means to shield these self-concepts from critical, self-threatening information.

In terms of the three aspects of epistemic beliefs, individuals with high levels in D are expected to trust their intuition when evaluating the accuracy of any information, implying a strong Faith in Intuition for Facts. As evidence bears the risk of contradicting one's worldview, we expect a negative association between D and Need for Evidence. The disregard of evidence can be supported by endorsing the idea that facts, including scientific evidence, are partially or completely constructed by society which allows for multiple perspectives on what is to be regarded as true (Kata, 2012) . Accordingly, we expect that individuals with high levels in D tend to hold the belief that Truth is Political. To sum up, we argue that a pronounced Dark Factor of Personality should be linked to post-truth epistemic beliefs.

We assume that D and post-truth epistemic beliefs are associated with specific cognitions and behavior in response to societal and political phenomena such as the handling of the COVID-19 pandemic. Conspiracy theories attribute the actual cause of an event to the intrigues of several powerful actors who are working towards a common goal that is contrary to the interests of large sections of the population . Conspiracy theories about COVID-19 range from downplaying its danger while suspecting others to profit from exaggerating the severity of the disease, to explicitly assuming that malevolent forces spread COVID-19 as a bioweapon (e.g., Imhoff & Lamberty, 2020) . The central idea of conspiracy theories, namely powerful actors secretly working towards a common goal against the will of the majority of the people, fits within the self-serving worldview of individuals high in D (Moshagen et al., 2018) . Ruthlessly prioritizing one's individual utility as the core motif of individuals with high levels in D, could well lead to the assumption that others think and behave similarly. Accordingly, Machiavellianism has been linked to a tendency to believe in conspiracy theories and increased willingness to conspire (Douglas & Sutton, 2011) . More recently, Machiavellianism, narcissism, and psychopathy were found to be associated with the endorsement of generic (Kay, 2021) and COVID-19 specific conspiracy theories (Ahadzadeh et al., 2021) . Following our line of argument, individuals high in D should trust their intuition when confronted with conspiratorial ideas and refrain from relying on evidence. The idea that what is regarded as "true" is dependent on politics and society, could further enhance conspiratorial thinking (Garret & Weeks, 2017) .

Compliance with countermeasures against COVID-19, including handwashing, wearing hygienic face masks and social distancing, is highly dependent on trust in government and science (Plohl & Musil, 2021) . As described above, individuals high in D should only rely on evidence and hence trust scientific recommendations, if it suits their agenda. In the context of COVID-19, however, this seems highly unlikely, as countermeasures rely on the engagement in prosocial behavior (Anderson et al., 2020; Han et al., 2021) , which is in stark contrast to the very definition of D (Moshagen et al., 2018) . Research has linked D as well as individual dark traits such as Machiavellianism, narcissism and psychopathy to less protective behavior against COVID-19 (Blagov, 2021; Nowak et al., 2020; Ścigała et al., 2021; Zettler et al., 2021) . Again, following our line of argument, individuals high in D should rely on post-truth epistemic beliefs to neglect scientific evidence regarding COVID-19 and devalue it, for example by trusting their intuition about its accuracy. They could also assume that recommendations aiming at the implementation of countermeasures are merely the result of one of multiple (scientific) points of view.

Much of the available research has demonstrated that dark traits are more common in the right-wing political spectrum (Duspara & Greitemeyer, 2017; Jonason, 2015) and were linked to traditionally conservative stands on major societal topics (Arvan, 2013) . Not all studies, however, corroborated these relationships. It appears that these associations depend on various factors, such as the particular traits considered and the measures applied (see, e.g., Vize et al., 2018; Zeigler-Hill et al., 2020) . In terms of epistemic beliefs, both Faith in Intuition for Facts and Truth is Political were linked to conservatism (Garret & Weeks, 2017) . Additionally, conservatism has been shown to be associated with the endorsement of COVID-19 conspiracy theories as well as less protective behavior (e.g., de Bruin et al., 2020; Freeman et al., 2020) . In addition to these associations, political orientation could moderate the paths between D, epistemic beliefs, and both COVID-19 variables. More specifically, the joint occurrence of high scores on D and a right-wing political orientation could provide particularly fertile ground for strong inclinations to endorse post-truth epistemic beliefs, conspiracy ideation, and harmful behavior. For left-leaning individuals high in D, such inclinations could be weaker, as their social environment is likely more critical towards these stands, thereby reducing their utility. Moreover, post-truth epistemic beliefs could relate to the outcome variables in different ways, with stronger links for individuals with a more right-leaning political orientation.

Despite its theoretical plausibility, the connection between dark traits and epistemic beliefs has not been investigated yet and empirical evidence for the link between epistemic beliefs and the endorsement of conspiracy theories rests on one study (Garret & Weeks, 2017) . The latter authors showed that all three epistemic beliefs were associated with the score on a conspiracist ideation scale, but the results were somewhat mixed when relationships to conspiracy-related assertions about specific topics were examined (e.g., "Vaccines cause autism"). Although Machiavellianism and psychopathy have been linked to endorsing COVID-19 conspiracy theories (Hughes & Machan, 2021) and engaging in less protective behavior (Triberti et al., 2021) , research on the characteristic handling of evidence associated with dark traits is still missing. Additionally, we intend to corroborate as well as to extend prior research using a measure of the core of dark traits. Our studies are not only the first to investigate the link between dark traits and epistemic beliefs, but also to propose a model taking into account the complex interactions between dark traits, epistemic beliefs and political orientation as well as their collective effect on COVID-19-related cognitions and behavior (for an overview, see Figure 1 ). Four studies are presented: In Study 1, we investigated the relationship between D and epistemic beliefs, while in Studies 2-4 we included links to the endorsement of COVID-19 conspiracy theories and engagement in COVID-19 protective behavior, both inside and outside the United States over the course of the pandemic. We expected D to be positively associated with Faith in Intuition for Facts and Truth is Political and negatively associated with Need for Evidence (Studies 1-4). We further expected D to be positively associated with the endorsement of COVID-19 conspiracy theories and negatively associated with COVID-19 protective behavior (Studies 2-4). We hypothesized that these associations would be mediated by Faith in Intuition for Facts, Need for Evidence and Truth is Political, with Faith in Intuition for Facts and Truth is Political being positively and Need for Evidence being negatively associated with the endorsement of COVID-19 conspiracy theories. For COVID-19 protective behavior, we expected the reverse pattern of correlations (Studies 2-4).

In Studies 2-4, we also investigated the potential moderating role of political orientation on the associations described above. Following preliminary results (Study 2), we expected that the associations between D and the endorsement of COVID-19 conspiracy theories and engagement in COVID-19 protective behavior would increase with a more conservative political orientation. We expected the same pattern for the association between D and Truth is Political. Additionally, we expected both the association between D and Need for Evidence and the association between Need for Evidence and the endorsement of COVID-19 conspiracy theories to increase with a more liberal political orientation (Studies 3 and 4).

For all studies presented in this article, we report how we determined our sample size, all data exclusions, and all measures in the study, and we follow the Journal Article Reporting Standards (JARS; Kazak, 2018) . All data, analysis code, research materials are available at https://osf.io/ g3xkw/. All studies were preregistered (Study 1: https:// aspre dicted.org/6886q.pdf; Study 2: https://aspre dicted. org/3mp8c.pdf; Study 3: https://aspre dicted.org/44q7j. pdf; Study 4: https://aspre dicted.org/958ad.pdf).

In the first study, we explored the relationship between D and the epistemic beliefs subscales Faith in Intuition for Facts, Need for Evidence, and Truth is Political. A more detailed report of a confirmatory factor analysis can be found in the online Supporting Information (S1).

Our required sample size was based on a study by Wolf et al. (2013) who systematically varied major model properties to evaluate sample size requirements for commonly used structural equation models using Monte Carlo data simulation techniques. Note that, despite being an ancillary analysis in our case, confirmatory factor analysis demands greater sample sizes than the zeroorder correlations presented below. Based on their results for similar models, we aimed for a final minimum sample size of 300 participants, and we recruited 407 participants via Mechanical Turk to account for exclusions and paid 1 USD. We excluded 55 participants because they failed to respond to our control question appropriately ("This is a control question. Please do not select any of the 7 options below."; for details on the exclusion criteria and wordings of the control questions in all four studies see S5.3). Furthermore, 30 participants were excluded because they showed unreasonably low response times of less than 90 s and one participant because of an unreasonably high response time of more than 2700 s, indicating careless responding. The final sample consisted of 321 participants (M = 37.12, SD = 10.73, 20-78 years, 38% female). In terms of educational attainment, 38.3% had graduated from high school, 48.3% had a bachelor's degree, 9.0% had a master's degree and 1.6% had a Ph.D. or higher. The remaining 2.8% completed some high school or trade school. Regarding ethnicity, 75.7% stated they were White, 10.3% Black, 6.2% Hispanic, 5.3% Asian or Asian American, and 0.6% Native American.

Epistemic beliefs were assessed with a 12-item questionnaire by Garret and Weeks (2017) capturing the three subscales Faith in Intuition for Facts (e.g., "I trust my gut to tell what's true and what's not", α = .92), Need for Evidence (e.g., "Evidence is more important than whether something feels true", α = .84) and Truth is Political (e.g., "Facts depend on their political context", α = .92) with four items per subscale. Items are answered on a 7-point scale, ranging from strongly disagree (1) to strongly agree (7).

The Dark Factor of Personality was assessed with the D16 short version (Moshagen et al., 2020) . It consists of 16 items with a 7-point scale, ranging from strongly disagree (1) to strongly agree (7, e.g., "My own pleasure is all that matters", α = .90). Table 1 shows all zero-order correlations between the epistemic beliefs subscales and D as well as their means and standard deviations. As expected, Faith in Intuition for Facts and Truth is Political correlated positively with D, r = .27, p < .001 and r = .50, p < .001, respectively, and Need for Evidence was negatively correlated with D, r = −.20, p < .001. Study 1 provides first evidence for the link between D and the epistemic beliefs subscales. Indeed, the higher the participants' tendency to maximize their individual utility, the less they were inclined to commit to "the unforced force of the better argument" in Habermas' terms (Habermas, 1996, p. 305) . These results build the foundation for Studies 2-4 that investigated the effects of D and epistemic beliefs on the endorsement of COVID-19 conspiracy theories and the engagement in COVID-19 protective behavior.

The study was conducted on March 21, 2020, 10 days after the World Health Organization declared the COVID-19 outbreak a pandemic (World Health Organization 

Based on an analysis with G*Power (Faul et al., 2009 ), the required sample size for identifying an association of r = .15, with α = .05 and power = .90 is 462. Accordingly, we aimed for 550 participants to account for potential exclusions. Participants were recruited via Mechanical Turk and were paid 1.50 USD. In total, 550 participants completed the questionnaire. As we relied on U.S. participants, 56 participants were excluded because they either used a VPN/VPS or a proxy to mask their country and/ or failed to provide an adequate description of the study in English implying they are not native speakers or bots or careless responders. Additionally, we excluded 26 participants because they failed to respond to at least one of our control questions appropriately (see Table S5 .3). Furthermore, 13 participants were excluded because they showed unreasonably low response times of less than 120 s and two participants because of unreasonably high response times of more than 2700 s, indicating careless responding. 1 The final sample amounted to 453 participants (M = 40.37 years, SD = 12.23 years, 19-78 years, 42.4% female). 2 In terms of educational attainment, 29.6% had graduated from high school, 53.2% had a bachelor's degree, 13.7% had a master's degree and 1.1% had a Ph.D. or higher. The remaining 2.4% completed some high school or trade school. Regarding ethnicity, 79.5% stated they were White, 7.5% Black, 7.5% Asian or Asian American, 4.6% Hispanic, and 0.5% Native American.

In Study 2, we relied on the same measures for the epistemic beliefs and D that were used in Study 1. Again, reliabilities were excellent or good (Table S5. 2). Furthermore, we assessed participants' endorsement of COVID-19 conspiracy theories. The items were based on popular COVID-19 conspiracy theories circulating at the time of assessment and a prior instrument for assessing the endorsement of conspiracy theories regarding the Zika virus (Piltch-Loeb et al., 2019). On a 7-point scale, participants indicated the likelihood of six statements (e.g., "COVID-19 is a biological weapon originally developed by the Chinese government."). Options ranged from not at all likely (1) to extremely likely (7), α = .92.

The items for COVID-19 protective behavior were based on the COVID-19 pandemic behavior guidelines provided by the WHO (2020a). Participants indicated how much they complied with six statements referring to behaviors to slow the distribution of the coronavirus in the last three days (e.g., "In the last three days I have stayed home, unless required for my job, to buy groceries, or to help those in need."). A 7-point scale was provided,

(1) Dark factor of personality T A B L E 1 Study 1: Means, standard deviations, and zero-order correlations of the continuous variables options ranged from strongly disagree (1) to strongly agree (7), α = .78. Two additional items were used to measure the interference of social life during the past four weeks due to physical health or emotional problems (Hays et al., 1993) . We used the mean of both items as a control variable for additional analyses that involved the behavioral dependent variable (Supporting Information S2). Political orientation was assessed with a 7-point scale (M = 3.58, SD = 1.84). Options ranged from extremely left (1) to extremely right (7). Please note that our sample's distribution of political orientation was somewhat rightskewed. In Studies 2-4, we further assessed the percentages of participants tested (positive) for SARS-CoV-2 (see Table S5 .1). We included participants irrespective of their test results. The means, standard deviations, and zeroorder correlations of all focal variables are displayed in Table S2 .1.

We used PROCESS version 3.4.1 (Hayes, 2018) for our main analyses. While D served as the independent variable, endorsement of COVID-19 conspiracy theories and COVID-19 protective behavior served as dependent variables, respectively. The epistemic belief scales were entered as simultaneous mediators. Data for D and epistemic beliefs were standardized prior to all mediation and moderation analyses. In this parallel multiple mediator model, antecedent variable D was modeled as influencing one dependent variable (endorsement of COVID-19 conspiracy theories or COVID-19 protective behavior) directly as well as indirectly through the three mediators, with the condition that no mediator causally influences another (Hayes, 2018) .

The path coefficients, standard errors, and p-values are shown in Figure 2a . The first analysis yielded a total effect of D on the endorsement of COVID-19 conspiracy theories, B = .79, SEB = .06, 95%CI [.68; .90], p < .001, indicating that individuals high in D tend to endorse COVID-19 conspiracy theories. D was associated with all three epistemic beliefs in the expected directions. Also, in accordance with our predictions, Faith in Intuition for Facts and Truth is Political were associated with the endorsement of COVID-19 conspiracy theories. Need for Evidence was unrelated to the endorsement of COVID-19 conspiracy theories. We found a significant indirect effect for Faith in Intuition for Facts, B = .09, SEB = .02, 95%CI [.05; .13] as a mediator, as well as for Truth is Political, B = .20, SEB = .03, 95%CI [.14; .27]. Need for Evidence did not serve as a mediator, B = .03, SEB = .02, 95%CI [−.02; .08]. These results show that the association between D and the endorsement of COVID-19 conspiracy theories can be F I G U R E 2 Study 2: Main results of the parallel mediator models with endorsement of COVID-19 conspiracy theories (a) and COVID-19 protective behavior (b) as the dependent variables. Solid paths indicate significant associations (p < .05), dashed paths are non-significant partly explained by the epistemic beliefs held by individuals high in D.

The second parallel mediation analysis yielded a significant total effect of D on COVID-19 protective behavior, B = −.44, SEB = .04, 95%CI [−.52; −.36], p < .001, indicating that individuals with high levels in D showed less COVID-19 protective behavior. Furthermore, Faith in Intuition for Facts and Need for Evidence predicted COVID-19 protective behavior (Figure 2b ). Note that although we expected a negative association between Faith in Intuition for Facts and COVID-19 protective behavior, the association was positive. Truth is Political was unrelated to the dependent variable in the joint model. As expected, we found a significant indirect effect of D on COVID-19 protective behavior, mediated by Faith in Intuition for Political orientation also moderated several additional paths of our mediation models, which will be outlined in the following (see Figure S2 for graphical depictions, and Tables S2. .14], p = .911.

In sum, Study 2 supports our proposed model taking into account D, epistemic beliefs and political orientation as well as their interactive effect on COVID-19-related beliefs and behavior. We present evidence for the link between epistemic beliefs and COVID-19-related beliefs and behaviors. Our results build upon existing empirical evidence for the connection between dark traits and the endorsement of COVID-19 conspiracy theories and protective behavior. We find a tendency of individuals with high levels in D to endorse COVID-19 conspiracies as well as to neglect COVID-19 protective behavior. Our results show that the effect of D on the endorsement of COVID-19 conspiracy theories and the engagement in COVID-19 protective behavior can be explained by posttruth epistemic beliefs held by individuals high in D and that the strength of the associations increases with a more conservative political orientation.

Due to constantly and rapidly changing circumstances during the COVID-19 pandemic, the need to replicate initial findings appeared to be of extraordinary importance. Thus, in Study 3 we aimed for a replication of Study 2 using a larger sample size to corroborate the results at a later stage of the pandemic. Study 3 was conducted six months after Study 2, on October 14 and 15, 2020. At the time, there were almost eight million reported cases of COVID-19 infections in the USA and around 214,000 registered COVID-19-related deaths (WHO, 2020b).

Before and during Trump's presidency, political polarization in the USA had increased (Abramowitz, 2018) and COVID-19 had become a partisan issue when Study 3 was conducted (Druckman et al., 2021; Gollwitzer et al., 2020; Makridis & Rothwell, 2020) . The politicization of COVID-19 led to the crucial question whether individual differences in epistemic beliefs could still contribute to explaining the endorsement of COVID-19 conspiracy theories and engagement in protective behavior and whether the interactions with political orientation found in Study 2 could still be observed. Apart from minor adjustments to the scales used to assess F I G U R E 3 Graphical representation of the interaction between D and political orientation with COVID-19 conspiracy theories (a) and COVID-19 protective behavior (b) as criteria (Study 2). Higher scores in political orientation indicate a more right-leaning orientation. Semitransparent scatterplots represent single data points COVID-19 protective behavior (see Measures), our hypotheses and methods remained identical to those in Study 2.

We aimed for a substantially larger sample size than for our previous studies based on extant recommendations for powering replications and interaction effects (Giner-Sorolla, 2018; Simonsohn, 2015) . A sample of 1156 Mechanical Turk participants was invited, and 1113 participants completed the questionnaire and were paid 1.10 USD. We excluded 164 participants that either used a VPN/VPS or a proxy to mask their country of access and/or failed to provide an adequate description of the study in English implying they were not native speakers or bots or careless responders . We excluded another 14 participants because they failed to answer at least one of our control questions correctly (see Table S5 .3). Furthermore, one additional participant was excluded because of an unreasonably low response time of less than 120 s and eight participants because of unreasonably high response times of more than 2700 s. Participants were asked to state both their current age as well as their year of birth. Three participants were excluded because there was a mismatch between the two pieces of information, which was another indicator of careless responding . The final sample amounted to 923 participants (M = 39.43 years, SD = 11.64 years, 19-78 years, 44.9% female). In terms of educational attainment, 29.6% had graduated from high school, 51.1% had a bachelor's degree, 14.0% a master's degree, and 2.0% had a Ph.D or higher. The remaining 3.3% completed some high school or trade school. Regarding ethnicity, 79.2% stated they were White, 7.4% Black, 7.3% Asian or Asian American, 3.9% Hispanic, and 0.1% Native American.

In Study 3, we relied on the same measures for D, epistemic beliefs, the endorsement of COVID-19 conspiracy theories and political orientation that were used in Studies 1 and 2. Again, reliabilities were excellent or good (see Tables S5.2 and S3.1 for the descriptive statistics and zero-order correlations) and political orientation scores were somewhat right-skewed (M = 3.57; SD = 1.76). As the WHO advice for the public had been updated since Study 2 (WHO, 2020a), we changed our items for COVID-19 protective behavior slightly in accordance with the WHO advice. The means, standard deviations, and zero-order correlations of the focal variables are reported in Table S3 .1.

We performed the same analyses as in Study 2. Path coefficients, standard errors, and p-values are shown in Figure 4 . The first analysis yielded a total effect of D on the endorsement of COVID-19 conspiracy theories, B = .50, SEB = .04, 95%CI [.42; .57], p < .001, corroborating the result from Study 2. D was associated with Need for Evidence and Truth is Political in the expected directions, but D was not significantly correlated with Faith in Intuition for Facts. Furthermore, the epistemic beliefs were associated with the endorsement of COVID-19 conspiracy theories in the expected directions. We found a significant indirect effect of D on the endorsement of COVID-19 conspiracy the- .02], did not serve as a mediator. These results further support the assumption that the association between D and COVID-19 protective behavior can be explained by the epistemic beliefs held by individuals high in D.

As in Study 2, we performed two moderation analyses including political orientation as a moderator variable and the endorsement of COVID-19 conspiracy theories and COVID-19 protective behavior as dependent variables respectively. Political orientation predicted both the en- Study 3 largely corroborated our proposed model at a later stage of the pandemic at times of increased political polarization. We found support for the role of posttruth epistemic beliefs and respective links to D, to the endorsement of COVID-19 conspiracy theories, and to the engagement in COVID-19 protective behavior. Study 3 yielded an additional significant indirect effect of D on the endorsement of COVID-19 conspiracy theories, mediated by Need for Evidence that was not found in Study 2. We also found an additional significant indirect effect of D on COVID-19 protective behavior, mediated by Truth is Political. We did not replicate the indirect effect of D on the endorsement of COVID-19 conspiracy theories, mediated by Faith in Intuition for Facts (which involved an association that was reversed to what we had expected) as well as most of the moderating effects of political orientation. Nevertheless, the substantial main effects of political orientation on the dependent variables do speak for the pivotal role of political orientation as a contributing factor to the endorsement of COVID-19 conspiracy theories and neglect of COVID-19 protective behavior.

After having found largely consistent support for our proposed model focusing on the antecedents and consequences of post-truth epistemic beliefs over the course of the pandemic, we aimed to corroborate our results outside of the United States to provide support for their cross-cultural validity. Study 4 was conducted in Germany on December 22, 2020. At this time, around 1,500,000 people or 1.9% of the German population had reportedly been infected with SARS-CoV-2 and around 27,000 COVID-19-related deaths were registered for this country (WHO, 2020b). 3 There are certain differences between the US and Germany with regard to public trust in science and political polarization. In a study that was conducted in the years 2019 and 2020, 13% of the German population stated that they had little or no trust in science, whereas it was 21% in the USA (Funk et al., 2020) . In addition, in the USA the degree of polarization regarding trust in science between people self-identifying as politically left or right was far more pronounced (Funk et al., 2020) . In both countries, COVID-19 conspiracy theories are endorsed by a substantial part of the public (although the popularity of specific theories varies). In Germany, 17% of the population agreed to the notion that the government used COVID-19 as an excuse to restrict civil liberties (Infratest dimap, 2020), whereas 25% of U.S. citizens thought that the outbreak and dissemination of COVID-19 was planned (Schaeffer, 2020) . In contrast to the inconsistent and multi-faceted assessment of COVID-19 by the US government and members of the Republican party, there was a major consensus among German parties (including the then ruling conservative party, Christlich Demokratische Union, CDU) in favor of science-based protective measures against COVID-19, such as wearing face masks, lockdowns, and social distancing (e.g., Dean et al., 2020; Deutsche Welle, 2020) . Apart from minor adjustments to the scales used to assess the endorsement of COVID-19 conspiracy theories (see Measures), our hypotheses and methods remained unchanged.

We used G*Power and point biserial correlations as our basis for determining our sample size a priori. Given r = .20, α = .05 and power = .80, a sample size of 191 is required. 4 To add power for the interaction analyses, the required sample size was doubled, yielding 382 participants (Giner-Sorolla, 2018; Simonsohn, 2015) . We aimed for 550 participants to account for potential exclusions.

Participants were recruited online from the crowdworking site clickworker.de and paid 1.30 €.

In total, 550 participants were invited to participate in our study and 539 participants completed it. We excluded 11 participants because they failed to respond to at least one of our control questions appropriately (see Table S5 .3). Furthermore, three participants were excluded because of unreasonably low response times of less than 120 s and seven participants because of unreasonably high response times of more than 2700 s. Another four participants were excluded because they were under the age of 18. One participant did not indicate their political orientation and therefore had to be excluded. The final sample amounted to 513 participants (M = 37.54 years, SD = 12.23 years, 18-73 years, 40.7% female). In terms of educational attainment, 32.4% had graduated from high school, 43.5% had a bachelor's degree or master's degree. Another 24.0% completed some high school and one participant had no degree.

In Study 4, we relied on the same measures for D, the epistemic beliefs, COVID-19 protective behavior, and political orientation that were used in Studies 1-3. Reliabilities were good, close to excellent (Table S5 .2), the distribution of political orientation was somewhat right-skewed (M = 3.62, SD = .99). Prior research showed that the content of popular conspiracy theories in Germany differed somewhat from those in the USA. Thus, a set of items based on COVID-19 conspiracy theories circulating in Germany during the time of assessment was used (Imhoff & Lamberty, 2020) . For the assessment of D, we used the German version of the scale we had used in the previous studies (Bader et al., 2021) . All other scales were translated to German using the committee method. 5 Please refer to Table S4 .1 for descriptive statistics (including zero-order correlations) of all focal variables.

We performed the same analyses as in Studies 2 and 3. Path coefficients, standard errors, and p-values are shown in Figure 5 . The first analysis yielded a total effect of D on the endorsement of COVID-19 conspiracy theories, B = .44, SEB = .05, 95%CI [.35; .54], p < .001, corroborating the results of Studies 2 and 3. D was associated with Need for Evidence and Truth is Political in the expected directions, but not significantly linked to Faith in Intuition for Facts. Furthermore, Need for Evidence and Truth is Political were associated with the endorsement of COVID-19 conspiracy theories in the expected directions, while Faith in Intuition for Facts was not significantly associated with endorsing COVID-19 conspiracy theories. As expected, we found a significant indirect effect of D on the endorsement of COVID-19 conspiracy theories, mediated by Need for Evidence, B = .04; SEB = .02; 95%CI [.01; .07], as well as a significant indirect effect of D on the endorsement of COVID-19 conspiracy theories, mediated by Truth is Political, B = .24; SEB = .04; 95%CI [.17; .31]. Faith in Intuition for Facts did not serve as a mediator, B = −.001; SEB = .003; 95%CI [−.01; .004]. These results extend our empirical insights to Germany, corroborating the assumption that the association between D and the endorsement of COVID-19 conspiracy theories can be partly explained by post-truth epistemic beliefs held by individuals high in D.

The second parallel mediation analysis yielded a significant total effect of D on COVID-19 protective behavior, B = −.33, SEB = .03, 95%CI [−.39; −.26], p < .001, replicating the findings from Studies 2 and 3 that individuals with high levels in D tend to show less COVID-19 protective behavior. Furthermore, Need for Evidence and Truth is Political predicted COVID-19 protective behavior in the expected directions ( Figure 5b ). As in Studies 2 and 3, Faith in Intuition for Facts was positively associated with COVID-19 protective behavior whereas we had expected an effect in the opposite direction. As expected, we found .01]. These results further support the assumption that the association between D and COVID-19 protective behavior can be partly explained by epistemic beliefs held by individuals high in D.

As in Studies 2 and 3, we performed two moderation analyses, in which we additionally included political orientation as a moderator variable. The endorsement of COVID-19 conspiracy theories and COVID-19 protective behavior served as dependent variables. Political orientation significantly predicted the endorsement of COVID-19 conspiracy theories, B = .14, SEB = . In sum, Study 4 largely corroborated the results found in the USA in a German sample. This speaks to the crosscultural role of post-truth epistemic beliefs in shaping conspiracy ideation and behavior.

The complex challenges that our globalized, postmodern world faces, require evidence-based decision-making on a societal level, but also the individual willingness to follow "the unforced force of the better argument" (Habermas, 1996, p. 305) and to adjust one's behavior accordingly. However, not everyone shows this willingness: At least some people may deliberately choose to believe what they want to believe and to shield their opinions from the rational discourse. Our studies, which were conducted in the context of the COVID-19 pandemic, are the first to investigate both the antecedents and the consequences of these kinds of epistemic beliefs (Garret & Weeks, 2017). In brief, our four studies demonstrate that individuals high in D, that is, individuals who show a general tendency to maximize their personal utility, are more likely to hold a set of epistemic beliefs that views truth as being shaped by those in power while having a comparably low Need for Evidence and a high Faith in one's Intuition for Facts (a set we labelled post-truth epistemic beliefs). In turn, these epistemic beliefs predict the endorsement of COVID-19 conspiracy theories and protective behavior.

In Study 1, we demonstrated that individuals high in D tend to hold a set of post-truth epistemic beliefs, that is, are less inclined to commit to reasoning and argument based on evidence. These preliminary results were supported and extended in Studies 2-4. In addition to the association between D and epistemic beliefs already identified in Study 1, we found that epistemic beliefs predict the endorsement of COVID-19 conspiracy theories as well as protective behavior. While Studies 2 and 3 were conducted at two different points of the pandemic and thus demonstrate the stability of the results over time, Study 4 provided support for the cross-cultural validity of our results. Apart from these general findings, several aspects need to be emphasized. First, although we consistently found indirect effects of D mediated by epistemic beliefs on both dependent variables, the link between D and endorsing COVID-19 conspiracy theories was somewhat stronger than between D and COVID-19 protective behavior across all studies. This could be a result of regulations set in place that mandated social distancing, wearing masks in public transportation, and so on. Thus, even if post-truth epistemic beliefs led to an opposition of the rules set in place, the link to overt behavior could have been somewhat weakened by these policies.

Second, although the core idea of the proposed model was supported across all studies, there were some differences in terms of indirect effects between Study 2 on the one hand and Studies 3 and 4 on the other. The association between D and the endorsement of conspiracy theories was consistently mediated by Truth is Political. Whereas Faith in the Intuition for Facts served as a mediator in Study 2, indirect effects of mediator Need for Evidence were observed in both other studies. The association between D and engaging in COVID-19 protective behavior was consistently mediated by Need for Evidence. Whereas Faith in the Intuition for Facts served as a mediator in Study 2, indirect effects of mediator Truth is Political were observed in both other studies. In sum, the indirect effect pattern results of Study 3 are remarkably similar to the results of Study 4, which was conducted in a different country, but at a comparable point of the pandemic.

Third, deviating from our hypotheses, higher Faith in Intuition for Facts was associated with more protective behavior across all studies when considered in the joint model. When looking at the zero-order correlations, Faith in Intuition for Facts was not significantly correlated with protective behavior in Studies 2 and 3, and positively correlated in Study 4. This heterogeneity suggests that the way Faith in Intuition for Facts translates into protective behavior depends on various contextual factors. Possibly, the dissemination of scientific knowledge over the course of the pandemic has, on average, aligned judgments based on scientific evidence and individuals' intuitions, both speaking to protecting oneself against the virus.

As our studies demonstrate, political orientation plays a role in explaining the endorsement of COVID-19 conspiracy theories and the neglect of protective behavior beyond epistemic beliefs and D. Both the endorsement of COVID-19 conspiracy theories and reduced protective behavior were related to a more right-wing political orientation, which is in line with prior research (Kim & Kim, 2021; Miller, 2020) . Especially in the case of the US samples (Studies 2 and 3), this finding likely reflects that the Trump administration and the Republican party downplayed the danger of COVID-19 and the efficacy of countermeasures while focusing on economic issues (Haberman & Cooper, 2020; Smith, 2020) . Interestingly, the main effect of political orientation on COVID-19 conspiracy theories was weaker and the effect on protective behavior not significant in the German sample. It seems plausible to assume that this finding mirrors the fact that political polarization is less pronounced in Germany compared to the US (Boxell et al., 2020) .

Note also that the influence of political orientation changed between Studies 2 and 3. While we found a main effect of political orientation on endorsing COVID-19 conspiracy theories and protective behavior in both studies, we found clear moderating effects of political orientation in Study 2, but not in Study 3. In other words, political orientation did no longer interact with individual differences in Study 3. We hypothesize that the difference is an effect of an increased politicization of COVID-19 over the course of the pandemic (Kuchler, 2020; Marsh, 2020) : While holding certain beliefs and showing certain behavior with respect to COVID-19 was more of an individual matter in the beginning of the pandemic, it more and more became a matter of one's political affiliation over the course of time. This hypothesis is supported by the stronger main effect of political orientation on COVID-19 protective behavior in Study 3 compared to Study 2.

Although the present set of studies provides evidence for our model on the antecedents and consequences of post-truth epistemic beliefs, there are several important limitations to be considered. First, the present studies are cross-sectional and non-experimental, which makes it difficult to conclude causality. Although it seems theoretically far more plausible to assume that stable personality characteristics (i.e., the Dark Factor of Personality) influence epistemic beliefs, which in turn influence conspiratorial thinking and protective behavior than vice versa, more research is needed to establish clear causal links.

Second, our focus here was exclusively on post-truth epistemic beliefs as predictors of conspiratorial thinking and the willingness to engage in protective behavior in the context of COVID-19. This does not rule out, but rather complements alternative perspectives, such as work that has highlighted the role of individual differences in analytic thinking (e.g., Pennycook et al., 2021) or numeracy (Hutmacher et al., 2022) .

Third, it remains to be examined how epistemic beliefs shape the processing and dissemination of information pertinent to challenging societal topics. We assume that individual differences in epistemic beliefs can explain how individuals deal with misinformation, including the processing of misinformation indicators and the spreading of misinformation.

Fourth, the present studies were conducted in a very specific context: the COVID-19 pandemic. We assume that our basic model, the nexus between D, post-truth epistemic beliefs, and conspiratorial thinking and behavior holds for other key challenges our society faces. More specifically, we hypothesize that the current model will replicate in the field of climate change: Post-truth epistemic beliefs, fueled by D, nourish conspiracy ideation and behavior that stands in contrast to science-based recommendations.

Post-truth phenomena such as conspiracy theories and related behavior are widely considered to be a major threat to individual and societal prospering. We present consistent and cross-cultural evidence for the pivotal role of post-truth epistemic beliefs, rooted in the Dark Factor of Personality, in explaining the endorsement of COVID-19 conspiracy theories and non-adherence to behavioral recommendations throughout the pandemic.

Our research highlights that individuals' worldviews about how one can and should construct reality need to be taken into account when addressing major challenges to humankind.

We have no known conflicts of interest to disclose.

Jan Philipp Rudloff and Markus Appel developed the research ideas and designed the studies and Jan Philipp Rudloff analyzed the data. Jan Philipp Rudloff, Fabian Hutmacher and Markus Appel interpreted the data and drafted the manuscript. All authors approved the final version of the manuscript.

Based on the regulations for conducting psychological research in Germany, no formal IRB approval was required. The studies followed the ethical guidelines of the APA and the German Psychological Society (DGPs).

All data, analysis code, and research materials are available at https://osf.io/g3xkw/. All studies were preregistered (Study 1: https://aspre dicted.org/6886q.pdf; Study 2: https://aspre dicted.org/3mp8c.pdf; Study 3: https:// aspre dicted.org/44q7j.pdf; Study 4: https://aspre dicted. org/958ad.pdf).

Jan Philipp Rudloff https://orcid. org/0000-0003-0254-9251 Fabian Hutmacher https://orcid. org/0000-0002-0018-2559 Markus Appel https://orcid.org/0000-0003-4111-1308 ENDNOTES 1 Note that we adjusted our low response time exclusion criterion from 90 s (Study 1) to 120 s (Studies 2, 3, and 4), because the number of items was substantially larger in the latter three studies. Some scholars (e.g., Paas & Morren, 2018; Read et al., 2021) recommend the exclusion of particularly long response times and we followed this advice. In addition to the analyses reported in the main text, we performed the analyses without applying the upper exclusion criterion. All major results remain virtually the same (see Supporting Information S5). 2 The final sample size was a bit smaller than planned. Given α = .05, power = .90, and a sample size of 453, we were able to detect an association of r = .151, not r = .150 as originally planned (Faul et al., 2009) . 3 The infection rate is comparable to the infection rate of 2.4% of the U.S. population at the time of assessment of Study 3 (WHO, 2020b). However, there were only half as many reported COVID-19 related fatalities per 100,000 inhabitants in Germany (33 per 100,000 on December 22, 2020) than in the USA at the time of assessment of Study 3 (65 per 100,000 on October 15, 2020). 4 Note that in Study 2 we aimed for 90% power, as it was the first study in which we applied our newly developed model, 80% power appeared adequate for the subsequent studies. Furthermore, in Study 4 we relied on point biserial correlations as a basis for our sample size calculation because we also assessed a dichotomous dependent variable (the usage of a contact tracing app), but decided to move these analyses to the Online Supporting Information. Based on a Pearson correlation of r = .20, α = .05, and power = .80, the required sample size would have amounted to 193 participants. 5 Additionally, we expanded our research towards possible effects of D and epistemic beliefs on the use of contact tracing apps, more precisely the Corona-Warn-App. Given that at the time the study was conducted the utility of this app was questioned by government officials and many people who were following other behavioral advice (Kreder, 2020; Spiegel, 2020) , we are cautious to interpret the data. A detailed report of the results pertaining to the use of the Corona-Warn-App as the dependent variable can be found in the Online Supporting Information S4.

The great alignment: Race, party transformation, and the rise of Donald Trump

Social media skepticism and belief in conspiracy theories about COVID-19: The moderating role of the dark triad

The theory of planned behavior

How will country-based mitigation measures influence the course of the COVID-19 epidemic? The Lancet

A lot more bad news for conservatives, and a little bit of bad news for liberals? Moral judgments and the Dark Triad personality traits: A follow-up study

Measuring the Dark Core of Personality in German: Psychometric properties, measurement invariance, predictive validity, and self-other agreement

Adaptive and Dark Personality in the COVID-19 Pandemic: Predicting health-behavior endorsement and the appeal of public-health messages

Gathering, processing, and interpreting information about COVID-19

Cross-country trends in affective polarization (No. w26669)

The Colbert report

Descartes' error: Emotion, reason, and the human brain

Political polarization in US residents' COVID-19 risk perceptions, policy preferences, and protective behaviors

Merkel warns Germany is on the "thinnest ice" as Europe realizes social distancing is here to stay

Germany: Angela Merkel unveils new coronavirus measures. Deutsche Welle

The self-concepts of people with Dark Triad traits tend to be weaker, less clearly defined, and more state-related

Does it take one to know one? Endorsement of conspiracy theories is influenced by personal willingness to conspire

Affective polarization, local contexts and public opinion in America

The impact of Dark Tetrad traits on political orientation and extremism: An analysis in the course of a presidential election

Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses

Coronavirus conspiracy beliefs, mistrust, and compliance with government guidelines in England

Science and scientists held in high esteem across global publics

Epistemic beliefs' role in promoting misperceptions and conspiracist ideation

Driving a wedge between evidence and beliefs: How online ideological news exposure promotes political misperceptions

Powering your interaction. Approaching Significance

Partisan differences in physical distancing are linked to health outcomes during the COVID-19 pandemic

Trump called the Coronavirus "deadly" in private while minimizing its risks in public, book reveals

Between facts and norms: Contributions to a discourse theory of law and democracy

The social construction of what

Trust in government and its associations with health behaviour and prosocial behaviour during the COVID-19 pandemic

Introduction to mediation, moderation, and conditional process analysis: A regression-based approach

The rand 36-item health survey 1.0. Health Economics

Mass media flow and differential distribution of politically disputed beliefs: The belief gap hypothesis

The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learning

Does postmodernism really entail a disregard for the truth? Similarities and differences in postmodern and critical rationalist conceptualizations of truth, progress, and empirical research methods

Trump and vaccination: The effect of political identity, conspiracist ideation and presidential tweets on vaccine hesitancy

It's a conspiracy: Covid-19 conspiracies link to psychopathy, Machiavellianism and collective narcissism

The role of motivated science reception and numeracy in the context of the COVID-19 pandemic

Misinformation and worldviews in the post-truth information age: Commentary on Lewandowsky, Ecker, and Cook

A bioweapon or a hoax? The link between distinct conspiracy beliefs about the coronavirus disease (COVID-19) outbreak and pandemic behavior

Zwei Drittel sehen in Verschwörungstheorien eine wachsende Gefahr für die Demokratie

How "dark" personality traits and perceptions come together to predict racism in Australia

Valuing myself over others: The Dark Triad traits and moral and social values

Thinking, fast and slow. Farrar, Straus and Giroux

Anti-vaccine activists, Web 2.0, and the postmodern paradigm: An overview of tactics and tropes used online by the anti-vaccination movement. Vaccine

Actors of the most fiendish character: Explaining the associations between the Dark Tetrad and conspiracist ideation

Editorial: Journal article reporting standards

The shape of and solutions to the MTurk quality crisis

Searching for general model of conspiracy theories and its implication for public health policy: Analysis of the impacts of political, psychological, structural factors on conspiracy beliefs about the COVID-19 pandemic

So steht es international um die Corona-Warn-App [How the "Corona-Warn-App" is doing in international comparison

Motivated social cognition: Principles of the interface

Trump and Biden accuse each other of politicising Covid-19 vaccine

The development of epistemological understanding

The case for motivated reasoning

Beyond misinformation: Understanding and coping with the "Post-Truth" era

Genesis or evolution of gender differences? Worldview-based dilemmas in the processing of scientific information

The real cost of political polarization: Evidence from the COVID-19 Pandemic

How Donald Trump and Joe Biden have turned the hunt for a coronavirus vaccine into a political football

New York City region is now an epicenter of the Coronavirus pandemic

Psychological, political, and situational factors combine to boost COVID-19 conspiracy theory beliefs

The dark core of personality

Measuring the dark core of personality

The role of epistemic beliefs in self-regulated learning

Adaptive and maladaptive behavior during the COVID-19 pandemic: The roles of Dark Triad traits, collective arcissism, and health beliefs

Please do not answer if you are reading this: Respondent attention in online panels

Beliefs about COVID-19 in Canada, the United Kingdom, and the United States: A novel test of political polarization and motivated reasoning

Fighting COVID-19 misinformation on social media: Experimental evidence for a scalable accuracy-nudge intervention

Cross-sectional psychological and demographic associations of Zika knowledge and conspiracy beliefs before and after local Zika transmission

Modeling compliance with COVID-19 prevention guidelines: The critical role of trust in science

Racing the clock: Using response time as a proxy for attentiveness on self-administered surveys

A look at the Americans who believe there is some truth to the conspiracy theory that COVID-19 was planned

Science audiences, misinformation, and fake news

Effects of beliefs about the nature of knowledge on comprehension

Aversive personality and COVID-19: A first review and meta-analysis

Small telescopes: Detectability and the evaluation of replication results

Trump gives up on virus fight to focus on economic recovery and re-election. The Guardian

Diese Updates der Corona-Warn-App sind offenbar geplant [These Updates to the "Corona-Warn-App" are apparently planned

Power, politics, and paranoia: Why people are suspicious of their leaders

Analytic thinking reduces belief in conspiracy theories

Social distancing is the right thing to do: Dark Triad behavioral correlates in the COVID-19 quarantine

Differences among dark triad components: A meta-analytic investigation

Sample size requirements for structural equation models: An evaluation of power, bias, and solution propriety

advic e-for-public World Health Organization. (2020b). WHO Coronavirus (COVID-19) Dashboard

The darker angels of our nature: Do social worldviews mediate the associations that dark personality features have with ideological attitudes?

The role of personality in COVID-19-related perceptions, evaluations, and behaviors: Findings across five samples, nine traits, and 17 criteria