Conspiracy theories, particularly those related to vaccines, are a growing concern with negative implications for public health. This study addresses the role of scientific literacy, specifically health literacy, in shaping vaccine-specific conspiracy beliefs, exploring the competing deficit and contextual models. The deficit model posits that a lack of scientific knowledge directly leads to negative attitudes towards science. In contrast, the contextual model argues that social and cultural factors exert significant influence, either independently or by moderating the impact of knowledge on attitudes. This study focuses on the influence of health literacy, religiosity, and political identification as potential predictors of vaccination conspiracy beliefs and examines whether religiosity and political identification moderate the relationship between health literacy and belief in vaccine conspiracies. This research is crucial because understanding the factors that contribute to vaccine hesitancy is essential for developing effective public health interventions. The COVID-19 pandemic further highlighted the dangers of these beliefs and their societal impact, emphasizing the need for this research.

Previous research has shown a correlation between scientific literacy and positive attitudes towards science, although the relationship is not consistently strong or causal. The deficit model, while suggesting a direct link between knowledge and attitude, has been challenged by the contextual model which incorporates socio-cultural variables as stronger influences. Studies in the US have found links between political ideology (conservative/Republican) and lower acceptance of vaccination, as well as links between religiosity and vaccine hesitancy, although these results are not universal. Prior research also indicates that the relationship between literacy and science attitudes can be moderated by contextual factors such as religiosity and political identification, particularly in domains like embryonic stem cell research and climate change. However, there's limited research on these moderating effects in the context of vaccination conspiracy beliefs, especially outside of the US. This study aims to fill this gap by examining the main and moderating effects of health literacy, religiosity, and political identification in Croatia during the COVID-19 pandemic.

This study employed a quantitative approach using a survey administered to 729 Croatian university students from various faculties. The shorter version of the HLS-EU-Q health literacy scale (HLS-EU-Q6) was used to measure health literacy, showing a Cronbach's alpha of 0.76. Vaccination conspiracy beliefs were measured using a seven-item scale (Cronbach's alpha = 0.95) adapted from Shapiro et al. (2016). Religiosity and political identification were assessed on 1-10 point scales. Control variables included parental education, place of residence, gender, and field of study. Structural equation modeling (SEM) was employed to analyze the relationships between the variables, addressing measurement errors and allowing for simultaneous analysis of complex relationships. Two models were tested: Model 1 included the main effects of health literacy, religiosity, and political identification, while Model 2 added interaction terms between health literacy and the contextual variables. Regression multiple imputation handled the minimal missing data. The maximum likelihood estimation method was used. The study compared the student sample to a nationally representative sample regarding religiosity and political identification, finding the students to be slightly less religious and less right-wing.

Bivariate correlations showed a negative correlation between vaccination conspiracy beliefs and health literacy (r = -0.25), and positive correlations between conspiracy beliefs and religiosity (r = 0.27), and between conspiracy beliefs and right-wing political identification (r = 0.15). Model 1, assessing main effects, revealed that lower health literacy and higher religiosity significantly predicted higher vaccination conspiracy beliefs. Political identification was not a significant predictor. Model 2, including interaction terms, confirmed a significant moderating effect of religiosity. The negative interaction coefficient suggests that the negative effect of health literacy on conspiracy beliefs is stronger among more religious individuals. Interaction probing showed that at one standard deviation above the mean of religiosity, the effect of health literacy was significantly negative (-0.75), while at one standard deviation below the mean, it was not significant (-0.30). The moderating effect of political identification was not confirmed. These results confirmed hypotheses H1, H2, and H2a, but not H3 or H3a. The models showed acceptable fit indices (CMIN/df < 3; RMSEA < 0.05; CFI and TLI > 0.9).

The findings support both the deficit and contextual models. Lower health literacy contributes directly to increased belief in vaccination conspiracies (deficit model), aligning with previous research but not fully explaining the phenomenon. Religiosity significantly predicts increased belief in conspiracies and moderates the impact of health literacy (contextual model), showing that the relationship between knowledge and belief is not uniformly causal. The lack of a significant effect for political identification may be due to the specific socio-political context of Croatia during the pandemic, where political parties did not strongly polarize on vaccine issues, suggesting the influence is more driven by institutional dynamics than ideology. The counterintuitive moderating effect of religiosity, where higher religiosity amplified the negative effect of low health literacy on conspiracy beliefs, might indicate a ceiling effect or that skepticism, even among religious individuals, can be mitigated by increased scientific understanding. The role of trust in science as a mediator should be investigated further.

This study demonstrates the importance of both health literacy and contextual factors, like religiosity, in predicting vaccination conspiracy beliefs. While health literacy is crucial, it's not sufficient to fully explain the phenomenon. The socio-political context significantly shapes the relationship between these variables. Future research should investigate the mediating role of trust in science, explore the nuances of the interaction effects of religiosity and health literacy, and examine the specific role of institutional versus ideological factors in shaping attitudes toward vaccination across diverse contexts.

The main limitation is the use of a student sample, potentially limiting generalizability. Students may differ in religiosity and political identification compared to the general population. The use of self-reported health literacy might be confounded with other factors impacting vaccination attitudes. Future research should utilize larger, nationally representative samples and employ more nuanced measures of health literacy and conspiracy beliefs, particularly examining beliefs specific to COVID-19 vaccines to strengthen the study's validity and generalizability.