Pseudoscientific beliefs, alongside paranormal beliefs, conspiracy theories, and science denialism, are prevalent and potentially harmful. These epistemically unwarranted beliefs lack substantial evidence. Existing questionnaires often intermingle these belief types or include measures of scientific knowledge, confounding analysis. The Pseudoscience Endorsement Scale (PES), originally developed in Spanish, focuses solely on pseudoscientific beliefs, mitigating confounding variables. This study aimed to validate the English version of the PES and examine the relationship between pseudoscientific beliefs and other unwarranted beliefs. The hypothesis was that endorsement of pseudoscientific beliefs correlates with other unwarranted beliefs. Furthermore, the study investigated cognitive and sociodemographic correlates of pseudoscientific beliefs, exploring the relationship between formal education, scientific knowledge, reasoning styles, and gullibility. Previous studies show conflicting results on the relationship between education and pseudoscientific beliefs, some suggesting a positive correlation, while others report a negative association with scientific knowledge. The role of cognitive factors like intuition versus analytic thinking and gullibility in the endorsement of pseudoscientific beliefs also requires further investigation. The study expected that gullibility would positively predict and analytic thinking would negatively predict pseudoscientific belief endorsement, along with specific sociodemographic expectations based on previous research.

The introduction thoroughly reviews existing literature on different types of unwarranted beliefs, highlighting their prevalence and potential harm. It discusses the limitations of existing scales measuring these beliefs, emphasizing the need for a scale specifically targeting pseudoscientific beliefs without confounds. The literature review also explores the conflicting findings on the relationship between formal education, scientific knowledge, reasoning styles, and gullibility, and the endorsement of pseudoscientific beliefs. Studies showing associations between paranormal beliefs and lower education levels, and the more nuanced relationship between higher education and pseudoscientific beliefs (sometimes positive correlations) are noted. The conflicting evidence regarding cognitive styles (intuition vs. analysis) and gullibility is also reviewed. The influence of sociodemographic factors such as gender and socioeconomic status on pseudoscientific beliefs is also addressed.

The study employed a preregistered design with hypotheses and analyses documented on AsPredicted.org. A sample of 510 U.S. volunteers, recruited through Prolific, participated. The study utilized the English version of the PES, translated using back-translation methods. The scale contains 20 items measuring endorsement of pseudoscientific beliefs, rated on a 1-7 Likert scale. Additional questionnaires included the Revised Paranormal Beliefs Scale (RPBS), Science Denialism items from the Pseudoscientific Belief Scale (SD-PBS), the Generic Conspiracist Beliefs Scale (GCB), a bullshit detection task (measuring receptivity and sensitivity), the Science Literacy Knowledge Questionnaire (SLKQ), and the Cognitive Reflection Test (CRT). Sociodemographic data (age, sex, political ideology, years of schooling, socioeconomic status, and ethnicity) were also collected. Data analysis included psychometric analysis of the PES (reliability, factor analysis), correlational analyses (Kendall's tau due to non-normal distributions), and linear regression models predicting different types of unwarranted beliefs using cognitive measures and sociodemographic variables as predictors. The study was approved by the Universitat de Barcelona's ethics committee, and participants provided informed consent.

The initial psychometric analysis revealed very high internal consistency for the PES (ω = 0.92). However, factor analysis identified a one-component solution, and eight items with weak loadings were removed, leading to a shorter, more robust sPES (ω = 0.90). The sPES correlated positively with other measures of unwarranted beliefs (RPBS, SD-PBS, GCB), supporting the hypothesis of a common underlying basis. Correlational analyses revealed that sPES scores were positively associated with bullshit receptivity and motivational quote ratings, but negatively associated with bullshit sensitivity, scientific literacy (SLKQ), and CRT scores. Age and political ideology (right-wing/conservative) showed positive correlations with sPES scores, while years of schooling showed a weak negative correlation. No significant difference was found between men and women's sPES scores although men tended to score higher (although the Bayesian analysis favored the null hypothesis). Regression analyses, while not pre-registered, showed that bullshit receptivity and scientific literacy were the strongest predictors of pseudoscientific belief endorsement across all four types of unwarranted beliefs, with CRT showing weaker effects.

The study successfully validated the English version of the PES, resulting in a shorter, more reliable sPES. The positive correlations between pseudoscientific beliefs and other unwarranted belief types suggest a shared cognitive foundation. The findings highlight the importance of gullibility (bullshit receptivity) as a key predictor of pseudoscientific beliefs, and the protective role of scientific literacy and cognitive reflection (albeit weak in regression). The lack of strong associations with gender and socioeconomic status challenges previous findings, possibly due to the broader scope of the pseudoscientific belief measure. The positive correlations with age and right-wing political ideology require further investigation. The correlational nature of the study limits the ability to infer causality; future studies manipulating cognitive and educational variables are needed.

The study provides a validated short version of the Pseudoscience Endorsement Scale (sPES) and demonstrates the links between pseudoscientific beliefs and other unwarranted beliefs, cognitive factors, and sociodemographic variables. Gullibility and lack of scientific literacy are key factors, while the role of cognitive reflection needs further exploration. Future research should focus on interventions aimed at improving critical thinking skills and scientific literacy to combat the spread of pseudoscientific beliefs. Longitudinal studies exploring causal relationships between cognitive factors and belief systems would also be valuable.

The study's correlational design limits causal inferences. The use of self-report measures for cognitive styles might introduce biases. The sample, while representative of the U.S. population in terms of gender, might not fully capture the diversity of belief systems across different subgroups. The inclusion of regression models not pre-registered warrants caution in interpreting the relative importance of predictive variables. The weak effect of CRT in the regression analyses suggests the need to explore other measures of cognitive reflection.