The role of personality traits and online behavior in belief in fake news

The study addresses how individual differences—specifically personality traits, dark triad traits, and tendencies toward careless online behavior—relate to the ability to discern real from fake news. Motivated by the proliferation and public-health consequences of misinformation and conspiracy theories (e.g., during COVID-19), the authors aim to clarify which psychological factors and demographics predict fake news detection. They extend prior Big Five-based work by incorporating Honesty-Humility (from HEXACO) and the dark triad (Machiavellianism, narcissism, psychopathy). The key hypotheses were: (H1) lower Honesty-Humility and (H2) higher dark triad scores would predict decreased accuracy in discerning real vs fake news; (H3) more careless online behavior (greater online disinhibition, risky online behavior, higher engagement with strangers, lower suspicion) would predict poorer discernment; (H4) younger age and (H5) lower religiosity would predict better accuracy; and (H6) no gender differences were expected. The broader purpose is to inform media literacy interventions by highlighting risk factors rooted in personality and online behavior.

The literature links personality to misinformation susceptibility. Prior research (Calvillo et al., 2021) found higher Agreeableness, Conscientiousness, and Openness and lower Extraversion related to better fake/real news discrimination, and higher Agreeableness/Conscientiousness to more vetting of news before sharing (Sampat & Raj, 2022). Honesty-Humility, absent from the Big Five, predicts disinhibited and risky online behavior (D’Agata & Kwantes, 2020) and is theoretically related to intellectual humility, which reduces susceptibility to fake news and conspiracy beliefs and increases fact-checking (Bowes & Tasimi, 2022; Koetke et al., 2022, 2023). The dark triad traits (psychopathy, narcissism, Machiavellianism) have mixed associations with conspiracy beliefs across studies; they may reflect low Honesty-Humility (Hodson et al., 2018). Some work links dark triad traits to spreading unverified alarming news (Triberti et al., 2021) and to lower engagement in health-protective behaviors during COVID-19. Beyond personality, careless online behaviors—including online disinhibition, risky behaviors, openness to forming relationships with strangers, and lower suspicion—are linked to vulnerability to deception and social engineering (D’Agata & Kwantes, 2020; D’Agata et al., 2021). Demographic findings are mixed for gender, generally show higher susceptibility among more religious individuals, and suggest younger people may be more susceptible to fake news and conspiracies in some contexts.

Design: Correlational survey study with behavioral judgment task assessing fake vs real headline discernment. Participants: N=510 adults recruited via Qualtrics Panels from the US and Canada (age 18–80, M=39.8, SD=15.8). Gender: 315 female, 188 male, 4 non-binary. Education: 83.3% some post-secondary; employment: 60.6% employed. Inclusion: ages 18–80, US/Canada residence, English fluency. Exclusions: partial responses not recorded; failed random-response validity checks; extreme completion-time outliers or those <50% of median time removed. Ethics approval obtained. Measures: - Demographics: gender, age, education, employment; religiosity via Pennycook et al. (2016) religious belief scale (5-point Likert). - Online behavior: • Risky Online Behaviors (ROB; D’Agata & Kwantes, 2020): frequency count of risky behaviors (0–4 scale); α=0.94. • Online Disinhibition (OD; D’Agata & Kwantes, 2020): 20 items, 1–5 scale; α=0.90. • Openness to Form Online Relationships (OFOR; D’Agata et al., 2021): 6 items, 1–5 scale, two subscales: Engagement (α=0.84) and Suspicion (α=0.63). - Personality: • HEXACO-PI-R 100 (Lee & Ashton, 2018): Honesty-Humility, Emotionality, Extraversion, Agreeableness, Conscientiousness, Openness; subscale α’s=0.74–0.85. • Short Dark Triad (SD3; Jones & Paulhus, 2014): Machiavellianism, narcissism, psychopathy; α’s=0.75–0.83. - Headline Task (fake vs real detection): 24 headlines (12 real from mainstream outlets across political spectrum; 12 fake labeled false by Snopes), drawn from Peter et al. (2021), stratified by prior accuracy (high, ~50-50, low; 8 per bin, 4 real/4 fake). Each headline judged “real” or “fake.” Performance indexed via signal detection theory: sensitivity d′ (accuracy) and response bias c. Procedure: After demographics, participants completed the Headline Task (randomized order of headlines). Then OD, ROB, OFOR, HEXACO-PI-R, and SD3 were administered in randomized order. Analysis: Pearson correlations tested associations between d′/c and demographics, online behaviors, and personality. Bonferroni correction set α=0.004 for multiple comparisons; otherwise α=0.05. Paired t-test compared gender on d′.

- Demographics and accuracy (d′): Age was not significantly correlated with d′ (r=0.03, p=0.49). Religiosity was negatively correlated with d′ (r=-0.24, p<0.001), indicating lower religiosity associated with better discrimination. No significant gender difference in d′ (t(501) = -0.77, p=0.44). - Online behavior and d′ (Bonferroni α=0.004): Lower Suspicion associated with poorer accuracy (Suspicion r=0.16, p<0.001; higher suspicion relates to higher d′). Greater Engagement with strangers (r=-0.19, p<0.001), higher Online Disinhibition (r=-0.23, p<0.001), and more Risky Online Behaviors (r=-0.18, p<0.001) associated with lower d′. - Personality and d′: Lower Honesty-Humility (r=0.22, p<0.001; higher HH relates to higher d′), lower Openness (r=0.16, p<0.001), and lower Conscientiousness (r=0.20, p<0.001) associated with poorer accuracy. Dark triad traits associated with lower accuracy: Machiavellianism (r=-0.19, p<0.001), narcissism (r=-0.22, p<0.001), psychopathy (r=-0.25, p<0.001). - Response bias (c): Bias toward responding “true” (higher c) correlated with greater Online Disinhibition (r=0.23, p<0.001), more Risky Online Behaviors (r=0.28, p<0.001), greater Engagement (r=0.19, p<0.001), higher Machiavellianism (r=0.22, p<0.001), higher narcissism (r=0.21, p<0.001), and higher psychopathy (r=0.26, p<0.001). Lower Honesty-Humility related to a bias toward “true” (r=-0.21, p<0.001). Other HEXACO traits were not significantly related to c. - Descriptive task performance: Mean d′=0.27 (SD=0.55); mean c=0.03 (SD=0.53).

Findings indicate that both personality and online behavioral tendencies shape the accurate assessment of information veracity. Lower Conscientiousness, Openness, and Honesty-Humility were linked to poorer fake/real headline discrimination; individuals higher on dark triad traits (Machiavellianism, narcissism, psychopathy) showed reduced accuracy and a response bias toward judging headlines as true. Careless online behaviors—greater disinhibition, willingness to engage with strangers, more risky behaviors, and lower suspicion—were similarly associated with lower accuracy and a pro-true bias. These results partially replicate prior work linking Conscientiousness and Openness to better fake news detection and extend the literature by highlighting Honesty-Humility and the dark triad as important correlates. The response-bias pattern suggests that reduced accuracy may reflect a general tendency to accept information at face value. The authors also discuss demographic patterns as broadly consistent with prior research on religiosity and age, and emphasize that a trust-related theme underlies susceptibility: individuals who are less trustworthy (low Honesty-Humility, high dark triad) or more trusting of online information/strangers show greater acceptance of unverified claims. Implications include incorporating personality-informed self-awareness and critical evaluation strategies into media literacy interventions.

Personality traits—especially Honesty-Humility, Conscientiousness, and Openness—and careless online behaviors are associated with individuals’ ability to discern real from fake online information. The inclusion of Honesty-Humility and dark triad traits offers a novel theoretical contribution, suggesting that general trust tendencies and social aversiveness underpin susceptibility and a bias to accept headlines as true. The authors recommend media literacy programs that promote critical thinking, fact-checking habits, and self-awareness of personal risk profiles (e.g., personality-linked vulnerabilities). Future research should broaden the set of cognitive and motivational predictors (e.g., critical thinking, fact-checking willingness), leverage alternative (e.g., peer-report) measures, and employ longitudinal or experimental designs to clarify causal mechanisms and improve intervention targeting.

Effect sizes were small, typical for complex psychological phenomena. Measures relied on self-report, potentially affected by social desirability and inaccuracies, especially for dark triad and online behavior. The set of predictors was not exhaustive (e.g., critical thinking, willingness to fact-check not directly assessed). The correlational design limits causal inference; unmeasured variables may explain associations. Future work should use alternative/peer-report measures, survey communities with strong misinformation beliefs, and implement longitudinal or experimental designs; preregistration and replication are encouraged.