Correlates and antecedents of belief in CCCT¹ have been identified by psychologists, sociologists, and political scientists. Factors related to belief in CCCT include aspects such as demographic characteristics, as well as psychological and ideological influences. In a large-scale study with participants from the Pacific Northwest, Sarathchandra and Haltinner (2021b) identified several factors that differentiated CCCT believers from those who did not believe in such reports. They found that, compared to non-believers, CCCT believers tend to be male, older, more conservative, more religious, and have a higher level of education. The study also found that participants who believed in CCCT experienced fewer negative emotions about climate change (e.g., worry, dread, and sadness) compared to non-believers. Moreover, believers had lower levels of trust in climate scientists and the mainstream media but greater trust in fossil fuel companies than participants who did not believe in CCCT. Believers and non-believers also differed in the sources from which they obtained information about climate change. While believers relied more on weather forecasts and the Fox News channel (a conservative news outlet), non-believers retrieved their information from (climate) scientists and news outlets such as MSNBC, CNN, or NPR, which are considered mainstream or liberal-leaning (Sarathchandra and Haltinner, 2021b). Moreover, belief in CCCT is linked to other types of conspiracy ideation. Those who believe in CCCT not only exhibit a general tendency to explain events through conspiracy theories (i.e., “conspiracy mentality”; Bruder et al., 2013; Bertin et al., 2021) but they are also inclined to endorse other specific conspiracy theories (e.g., those related to COVID-19; Freeman et al., 2022). Interestingly, belief in specific conspiracy theories that do not primarily focus on climate change is negatively related to empirically warranted, non-conspiratorial beliefs about the existence, origin, and impacts of climate change (Pan et al., 2022).

Similarly, rejecting climate science is predicted by both climate-specific conspiracy beliefs and conspiracy beliefs that are unrelated to science (e.g., about JFK) (Lewandowsky et al., 2013). Van der Linden (2015) explored the causal link between belief in CCCT and acceptance of climate science through an experiment. Participants were assigned to one of three groups: a ‘conspiracy’ group that watched a video about CCCT, a ‘pro-climate’ group that viewed a video about taking action against climate change, and a control group that did not watch a video. Participants in the conspiracy condition expressed significantly greater agreement with the statement that climate change would be a hoax compared to those in the control group. Additionally, participants in the conspiracy condition believed there was less consensus on human-made climate change compared to both the control group and the pro-climate group. These findings highlight that the acceptance or rejection of climate science plays a role in support for CCCT.

Another factor that has been meta-analytically linked to the belief in climate science is the extent to which people trust scientists (Hornsey et al., 2016). Distrusting scientists is not only related to skepticism about climate change but also associated with the endorsement of CCCT, as shown by Sarathchandra and Haltinner (2021b). However, the relationship between distrust in scientists and conspiracy beliefs is not limited to the topic of climate change. In several correlational studies, distrust in science or scientists was related to COVID-19 conspiracy beliefs (Tonković et al., 2021; Chayinska et al., 2022; Vranic et al., 2022), generic conspiracy beliefs about malevolent actions of governments, secret organizations, and extraterrestrial contact (Fasce and Picó, 2019), and conspiracy mentality in general (Nera et al., 2022). Recognizing the multiple links between conspiracy beliefs and distrust in science and scientists, we aim to examine the factors and conditions affecting these relationships. More specifically, we focus on the relationship between distrust in climate science and scientists and belief in CCCT. Because educational research has established a link between EBs and trust in science within different domains (e.g., Strømsø et al., 2011; Schoor, 2023), we investigate whether EBs are also related to belief in CCCT and whether this relationship is mediated by trust in climate science.

EBs are beliefs about the nature of knowledge and knowing (Hofer and Pintrich, 1997). According to Hofer and Pintrich’s (1997) framework, the term “nature of knowledge” refers to the question “What is knowledge?” while “nature of knowing” is reflected in the question “How does one acquire knowledge?” and is, therefore, more procedural. Beliefs about the nature of knowledge consist of beliefs about the certainty of knowledge and beliefs about the simplicity of knowledge. Beliefs about the “certainty of knowledge” (Hofer and Pintrich, 1997, p. 119) span from the belief that knowledge is absolute and does not change to the belief that knowledge is provisional and may change over time. Beliefs about the “simplicity of knowledge” (Hofer and Pintrich, 1997, p. 120) range from the view that knowledge consists of a collection of more or less isolated facts to understanding it as strongly interrelated concepts that may depend on their context. Beliefs about the nature of knowing comprise beliefs about the source of knowledge and the justification for knowing. Beliefs about the acquisition of knowledge (“source of knowledge,” Hofer and Pintrich, 1997, p. 120) range from the belief that knowledge is received from external authorities such as experts or scientists to the belief that knowledge is actively constructed through interaction with others. Beliefs about the justification basis of knowledge (“justification for knowing,” Hofer and Pintrich, 1997, p. 120) range from justifying claims relying on personal observation, authority, or intuition to justifying claims employing inquiry rules, evaluating different sources, and integrating them. EBs can change over time and get more sophisticated as a person’s level of education increases (King and Kitchener, 2001). Kuhn et al. (2000) propose a developmental perspective that describes the progression from rather low-level EBs to sophisticated ones. They argue that at the lowest levels, the realist and absolutist level, knowledge is seen as objective, certain, either right or wrong, and located in the external world. At the multiplist level, however, knowledge is seen as a product of human cognition and as provisional. At this stage, assertions are seen as subjective opinions, and all opinions hold equal value. In the last stage, the evaluativist level, knowledge is also seen as uncertain and created by humans. However, claims are viewed as “judgments” (Kuhn et al., 2000, p. 311), which can be verified and substantiated by evidence and arguments. Nevertheless, the degree of sophistication of beliefs about the nature of knowledge and knowing depends on the context, that is specific scientific topics and domains (Elby and Hammer, 2001; Sinatra et al., 2014). Although the notion of “tentative knowledge,” for example, is considered sophisticated (Kuhn et al., 2000), Sinatra et al. (2014) argue that it would not “be beneficial to an understanding of science to think that there is an ever-present, similar degree of uncertainty on every topic” (p. 127). Using climate change as an example, they explain that it is reasonable to recognize the tentativeness of scientific knowledge in general while also acknowledging the scientific consensus on human-induced climate change². This context-specificity of EBs also has implications for the measurement of EBs (Bråten et al., 2009). A measure that takes context-specificity into account was introduced by Peter et al. (2016), who argued that sophisticated EBs are manifested in low endorsement of both generalized (i.e., not context-specific) absolutist and generalized multiplist statements.

EBs can influence how people deal with (every day) scientific information. Sinatra et al. (2014) suggest that EBs play a crucial role in assessing scientific claims, particularly when individuals encounter contradictory information and must evaluate it carefully. Moreover, EBs come into play when people have to incorporate new knowledge into existing knowledge or need to integrate knowledge from different sources. We propose that EBs might play a role when people are confronted with conspiratorial claims about scientific issues like COVID-19, genetically modified food, or climate change. This is because EBs have been related to common correlates of conspiracy beliefs, such as the need for closure (DeBacker and Crowson, 2006; Peter et al., 2016), the need for cognition (Nussbaum and Bendixen, 2003), and critical thinking (Muis et al., 2021). Also, EBs predict conspiracy beliefs. In an early study, Garrett and Weeks (2017) used another conceptualization of EBs with the dimensions “faith in intuition for facts,” “need for evidence,” and “truth is political.”³ They found that relying on one’s gut feeling when evaluating facts predicts stronger conspiracy beliefs (e.g., about the Moon landing or the assassination of JFK), as does the belief that facts and truth are determined by those in power. In contrast, valuing evidence predicts lower conspiracy beliefs. Moreover, participants with a higher need for evidence were more likely to believe that climate change is human-made, while participants who thought that the truth would be political were more likely to disbelieve this. The relationships between faith in intuition, the need for evidence, and the idea that truth would be political were replicated with respect to COVID-19-related conspiracy beliefs (Rudloff et al., 2022). Serrano et al. (2023) also explored the relationship between EBs and COVID-19-related conspiracy beliefs, following Peter et al.’s (2016) conceptualization of EBs. They showed that college students’ beliefs that knowledge is subjective, uncertain, and constructed by their own reasoning (i.e., multiplist beliefs) were positively linked to conspiracy beliefs about COVID-19. This relationship was mediated by trust in science. Multiplist beliefs were negatively related to trust in science, which, again, was negatively related to COVID-19 conspiracy beliefs. Consistent with Serrano et al. (2023), we expect that EBs about climate change will predict believing CCCT.

Similar to the work by Serrano et al. (2023), other studies have also observed a connection between EBs and trust in science and scientists. Participants who believed knowledge was uncertain and preliminary and who relied on their intuition to justify knowledge claims were less likely to trust scientists and science in general (Schoor, 2023). In the same study, Schoor (2023) found that participants who justified knowledge claims based on authorities such as experts were more likely to trust in science and scientists. These findings are complemented by a study by Strømsø et al. (2011), which found that participants who believed that knowledge about climate change primarily originates from their own thinking and interpretations – rather than relying on expert statements – were less likely to trust a popular science text about climate change. In comparison, participants who believed knowledge about climate change should be justified by thinking critically and comparing multiple sources were more likely to trust this text. Moreover, faith in intuition when judging claims is negatively related to believing information from scientific and liberal sources compared to conservative ones (Butterfuss et al., 2020). Although their study did not find a direct link between EBs and trust in these sources, it suggested that EBs might be connected to political ideology. Building on these insights, we anticipate that EBs about climate change will be linked to distrust in climate science and will seek to examine whether political ideology also plays a role in the relationship between EBs and distrust in science.

We test the following hypotheses in one correlational study and one experimental study:

EBs about climate change will correlate with and predict belief in CCCT (H1). More specifically, we expect that certainty beliefs and acquisition beliefs positively correlate with and predict belief in CCCT. We expect that justification-based beliefs negatively correlate with and predict belief in CCCT. We also expect that simplicity beliefs correlate with and predict belief in CCCT.

Trust in climate science will mediate the relationship between EBs about climate change and belief in CCCT (H2).

Political ideology will moderate the relationship between EBs about climate change and trust in climate science (H3).

The correlations between all variables can be found in Table 1.

In Study 1, we examined the relationships between EBs about climate change, distrust in climate science, and belief in CCCT. To measure EBs about climate change, we developed an ad hoc scale to measure the four dimensions of EBs by Hofer and Pintrich (1997: certainty, simplicity, source, justification of knowledge). However, a confirmatory factor analysis indicated that the four-factor model did not adequately fit the data and that a five-factor structure provided a significantly better fit. This five-factor structure consisted of the following dimensions: beliefs about the certainty of knowledge, structure of knowledge, the self as a source of knowledge, experts as a source of knowledge, and the subjective justification of knowledge. While the dimensions of certainty and structure (“simplicity” in the Hofer and Pintrich (1997) framework) remained unchanged, the source dimension was split up into two dimensions (i.e., the self vs. experts as a source of knowledge), and the justification dimension only comprised subjective justification. By distinguishing between self and experts as sources of knowledge, we recognize that individuals can simultaneously engage in personal research and rely on scientific authorities, allowing for a nuanced assessment of EBs that can be used to reveal effects that were previously less apparent. Moreover, we think that beliefs about the justification of knowledge are strongly connected to beliefs about the source of knowledge (Greene et al., 2008). Therefore, we believe that beliefs about experts as a source of knowledge about climate change may as well overlap with the justification by authority (i.e., experts or scientists) dimension that Greene et al. (2008) differentiated from personal justification.

Study 1 suggests that different views about climate change knowledge predict greater or lesser belief in CCCT. More specifically, participants who believed knowledge about climate change is provisional or who believed climate change knowledge is best grounded in gut feeling and personal observation endorsed stronger belief in CCCT. Conversely, participants who believed knowledge about climate change is interrelated and complex, as well as those believing experts and science are the main sources of knowledge about climate change, endorsed lower belief in CCCT. Distrust in climate science mediated all the relationships between the different EBs’ subscales and belief in CCCT. This suggests that different views about climate change knowledge affect one’s (dis-)trust in climate science, which, again, affects whether someone believes conspiracy theories about climate change. Contrary to our expectations, political ideology only affected the strength of the relationship between believing that experts were the main source of knowledge about climate change and distrust in climate science. Strikingly, the positive relationships between the EBs’ subscales and distrust in climate science did not depend on political ideology. This indicates that for these individuals, skepticism toward climate science may arise from perceptions of uncertainty and reliance on personal observations rather than from their political beliefs. Another explanation may be the limited variability of the political ideology scores that lay between 1.78 and 4.14 on a 7-point scale (M = 2.96, SD = 1.18), which may have led to a reduced power for detecting moderation (Memon et al., 2019). Finally, the new dimension, “self as a source of knowledge,” was neither related to distrust in science nor to belief in CCCT, potentially due to the low reliability of the subscale. Taken together, our results corroborate prior research that linked different conceptualizations of EBs to belief in conspiracy theories (Garrett and Weeks, 2017; Rudloff et al., 2022; Serrano et al., 2023). Similar to Serrano et al. (2023), we found that distrust in science mediates the link between EBs and conspiracy beliefs. By examining a different conspiracy narrative and using an alternative conceptualization of EBs, we found initial evidence for the robustness of this relationship. Moreover, our findings contribute to the evidence documenting the links between EBs and trust in science (Strømsø et al., 2011; Schoor, 2023) as well as between trust in science and conspiracy beliefs (Fasce and Picó, 2019; Tonković et al., 2021).

One limitation of our study was the partly low reliability of our ad hoc EBs measure, potentially leading to an underestimation of the relationships between EBs and belief in CCCT or distrust in climate science (Schmitt, 1996). Yet, despite the low reliability of the measure, we were able to establish moderate relationships between the EBs’ subscales, distrust in climate science, and climate-related conspiracy beliefs. One exception to this is the scale that was intended to measure the dimension of “Self as a source of knowledge.” It had poor reliability coefficients and only correlated with subjective justification beliefs. Moreover, our study was correlational, which is why we cannot draw causal conclusions. Therefore, we conducted a second, experimental study.

Descriptive statistics can be found in Tables 2, 3, and the correlations between the variables can be found in Table 4.

In Study 2, we attempted to establish a causal link between EBs and climate conspiracy beliefs. To achieve this, we randomly assigned participants to one of three conditions: a “naïve” EBs condition, a “sophisticated” EBs condition, and a control condition. However, the experimental manipulation only affected certainty beliefs, with participants in the “naïve” condition having significantly lower certainty beliefs than those in the “sophisticated” condition. All other EBs variables, and most importantly, belief in CCCT and distrust in climate science, were not affected by our experimental manipulation. Thus, the experimental groups did not differ in predicting distrust in climate science or belief in CCCT, and the predicted (moderated) mediation was not present. On the correlational level, however, we could replicate all relationships between the EBs’ subscales and distrust in climate science as well as belief in CCCT.

The experimental manipulation may have failed because participants could skip the page containing the text. The rather short reading times suggest this explanation. We think that some of the participants may have only read the first sentence, if at all, which begins as follows: “Knowledge about climate change is established and does not simply change” (naïve beliefs condition) or for the sophisticated beliefs condition: “Knowledge about climate change is still provisional and is constantly evolving.” The first sentence aimed to manipulate participants’ beliefs about certainty. Thus, the possibility that participants only read the first sentence or the presence of a primacy effect (Sullivan, 2019) might explain the partial success of the experimental manipulation. However, it is also possible that the manipulation texts were insufficient to induce changes in participants’ EBs. To address this issue, future studies could employ alternative manipulation strategies, such as targeting each dimension of EBs individually to make the manipulations more effective. Another explanation for the absence of an effect on belief in CCCT or distrust in climate science could be the fact that the study was slightly underpowered.

Despite the insights gained from this study, it is important to acknowledge its limitations. First, our data were collected only in Germany, which restricts our findings only to a WEIRD (Western, Educated, Industrialized, Rich, and Democratic; Henrich et al., 2010) population. This has theoretical implications, for example, regarding trust in science. While trust in scientists is higher among the political left overall, this relationship varies between countries and is not evident or even reversed in some countries (Cologna et al., 2024; preprint). Therefore, the moderating effect of political ideology that we found on the relationship between EBs and trust in science might differ when examined in other countries. With regard to the impact of political ideology on distrust, it is important to note that in both samples, participants’ political views were predominantly centered or left-leaning. In a sample with more right-leaning participants, the moderating effect of political ideology might be more pronounced, particularly among European or North American individuals. Moreover, we used a unidimensional measure of political ideology ranging from left to right. Such unidimensional conceptualizations of political ideology have been challenged because people can interpret the meaning of left and right differently (Bauer et al., 2017). At the same time, how these dimensions are understood differs between countries or cultures (Zechmeister, 2015; Hsiao et al., 2017). These variations in associations with left and right can affect the validity of political ideology measurements (Bauer et al., 2017). An alternative to this unidimensional approach is the use of multidimensional measures that include cultural and economic dimensions (Feldman and Johnston, 2014; Alves and Porto, 2022). However, we chose the ALLBUS left–right self-placement measure from the German General Social Survey, as our data were collected in Germany, where left and right are commonly used terms that simplify the concept of political ideology. Although the generalizability of our findings beyond German society is limited, it is noteworthy that a significant portion of our data was collected from non-student participants in the second study (with 76% students in Study 1 and 39% students in Study 2). This enables some generalization beyond the university student population, extending our findings to a broader group, although the sample remains non-representative.

Finally, we want to clarify that, after careful consideration, we have decided to make some deviations from our preregistration. This includes modifications to the first hypothesis and changes to the sample size in the second study. While it is common for researchers in psychological science to deviate from their preregistrations (Claesen et al., 2021; Willroth and Atherton, 2024), we believe that being transparent about these changes is essential for maintaining the credibility and replicability of the research. Therefore, the details of the deviations, along with their justifications and potential implications, are documented in our repository on the Open Science Framework (OSF; Foster and Deardorff, 2017).

The study highlights the crucial role of EBs and trust in science in shaping belief in climate change conspiracy theories. Theoretically, it deepens our understanding of the cognitive and trust-based mechanisms underlying conspiracy thinking. Practically, it suggests strategies for improving science communication, education, and policy to foster EBs that reduce distrust in climate science and conspiracy beliefs.

Effective science communication may help shape EBs in real-world settings by promoting beliefs that support the scientific approach while diminishing those that challenge it. To counteract perceptions of climate knowledge as uncertain, scientists and communicators should avoid emphasizing uncertainty in their messaging—even though uncertainty is undoubtedly a core value of science (Clegg, 2010). As the perceived certainty of climate change messages influences their perceived plausibility (Lombardi et al., 2014), communicating the stability of findings about climate change may also affect EBs about the certainty of climate change knowledge. Correspondingly, it is necessary to avoid false-balance media reports (i.e., balancing scientific consensus with contrarian voices in media reports) as they can lower the perception of a scientific consensus about climate change (Cook et al., 2017).

In addition to science communication, another cornerstone for changing EBs is education, particularly in secondary education. To address the reliance on personal observation in justifying climate change knowledge, educational initiatives can incorporate simple, relatable experiments that demonstrate concepts like the greenhouse effect and the water cycle. These hands-on activities connect scientific principles to everyday experiences, satisfying the need for personal observation while making scientific explanations more accessible and appealing. This idea is supported by a study by Xie et al. (2023), who found that participating in informal scientific activities is related to EBs that value scientific methods to justify knowledge. Moreover, personal justification can also be reduced through interventions that involve critical engagement with conflicting information from multiple sources about a scientific topic (Ferguson and Bråten, 2013). At the same time, such interventions can heighten the justification through science. Another approach to strengthening or supporting the development of EBs about science as a source of knowledge could involve educational initiatives aimed at enhancing science literacy. Science literacy can be defined as knowledge of scientific concepts, theories, and methods used to prove scientific hypotheses empirically. It is, therefore, associated with the ability to interpret and evaluate (pseudo-) scientific evidence (OECD, 2017). There is ample evidence that science literacy can be actively improved, for example, through participation in science courses (Surpless et al., 2014; Mahler et al., 2021) or through hands-on, problem-based learning in middle school classrooms (Lawless and Brown, 2015; Vieira and Tenreiro-Vieira, 2016). EBs are also positively related to science literacy (Wang et al., 2022), which is why we believe there is potential to foster pro-science EBs by enhancing science literacy. Finally, education predicts reduced belief in simple solutions to complex problems (Van Prooijen et al., 2015; van Prooijen, 2017) – which again is linked to conspiracy belief. Therefore, we argue that education can be a solution to enhance EBs about the complex structure of knowledge. On the theoretical level, one can, therefore, argue that EBs may also be a driving factor for the relationship between education and conspiracy theories (van Prooijen, 2017; Muscat-Inglott, 2023).

Our findings, combined with prior research on strategies to influence EBs, underscore the need for education-focused policies. These include teacher training, adult education programs, initiatives to enhance science literacy, and support for museums or informal learning environments, as previously recommended in a report for the European Parliament’s Committee on Culture and Education (Siarova et al., 2019).