The Role of Conspiracist Ideation and Worldviews in Predicting Rejection of Science

Trust in science among American conservatives has been declining since the nineteen seventies. Among liberals, that trust has remained stable. This single historical asymmetry is the starting point for a study by Stephan Lewandowsky, Gilles Gignac, and Klaus Oberauer. Their central argument is that when you look closely, science rejection turns out not to be a singular phenomenon. To test that idea, they chose three scientific domains: climate change, genetically modified foods, and vaccinations. This choice was deliberate. Climate science carries obvious regulatory implications—accepting it means accepting that emissions policy is necessary. Genetically modified foods are widely debated but do not threaten the economic status quo in the same way. Vaccinations invoke both public health regulation and personal liberty. When you put those three side by side, you have a natural experiment for separating the drivers of rejection. Lewandowsky and colleagues tested three candidate predictors. First, self-reported political conservatism. Second, endorsement of a free market worldview—a preference for minimal regulation and market solutions. These two were correlated, with a correlation coefficient of zero point eight one, but the authors treated them as distinct constructs. Third, and most importantly, conspiracist ideation—not belief in any single conspiracy theory, but a general cognitive tendency to endorse conspiracy explanations across a wide range of topics. To measure all of this, they ran a propensity-weighted internet panel survey of one thousand and one U.S. residents. Propensity weighting corrects for who tends to participate in online panels versus who makes up the actual population. It ensures that researchers are not just measuring the opinions of unusually engaged internet users. The thirty-nine item survey used a five-point agreement scale, with five items each for climate, genetically modified foods, and vaccines, as well as separate scales for free market endorsement, conservatism, and conspiracist ideation. They then modeled everything using structural equation modeling, or SEM. The key advantage of SEM is that it allows you to isolate each predictor's independent contribution while holding the others constant. That matters because conservatism and free market worldview are so tightly correlated that looking at either one alone can give a misleading picture. The model fit the data very well with a comparative fit index of zero point ninety-nine and a root mean square error of approximation of zero point zero four five, well within the accepted thresholds for an excellent fit. Now for the results. On climate science, the worldview predictors were powerful. In the full structural equation model, free market endorsement carried a standardized weight of negative zero point thirty-two, and conservatism carried negative zero point forty-nine. Both were strongly associated with rejecting the science of human-caused warming. When modeled separately, those effects grew even larger—negative zero point seventy for free market endorsement alone and negative zero point seventy-six for conservatism alone. The authors interpret this through a regulatory threat lens: people whose worldview depends on minimal government intervention resist scientific findings that logically call for regulation. The threat isn't to facts; it is to values. There is also evidence that this resistance can be partially addressed. Lewandowsky and colleagues cite experimental work showing that emphasizing the breadth of scientific consensus increases endorsement of climate change, and that consensus messaging was particularly effective for people whose worldview would otherwise predispose them toward rejection. Perceived consensus turned out to be the strongest predictor of climate acceptance among Republicans specifically. That's an important practical signal, and we'll return to it. The result for vaccines was more complicated. In the full model, free market endorsement predicted greater rejection of vaccinations, while conservatism predicted greater acceptance—effects pointing in opposite directions. Bivariate correlations between each worldview variable and vaccine acceptance were small and non-significant. What the structural equation modeling revealed was a suppressor effect: Because conservatism and free market endorsement are so highly correlated, their opposing influences on vaccine attitudes cancel out in simple analyses. Only the multivariate model can pull them apart. When free market endorsement was removed from the model, the conservatism link to vaccination became non-significant. When conservatism was removed, the free market link remained significant but small, around negative zero point fourteen. Genetically modified foods showed essentially nothing. The authors tested whether the worldview predictors contributed anything to genetically modified rejection and found they could set those paths to zero without any significant loss of model fit—a chi-square difference of four point eight with two degrees of freedom and a p-value of approximately zero point ten. Political worldview simply does not predict who opposes genetically modified foods. So, conservatism and free market endorsement are strong and specific predictors—powerful for climate, weak and opposing for vaccines, and absent for genetically modified foods. That pattern alone tells you that science rejection is not a single political phenomenon. But then there's the third predictor, and this is where the story gets more interesting. Conspiracist ideation predicts rejection across all three domains. Climate science, genetically modified foods, and vaccinations — the same general tendency to endorse conspiracy theories is associated with lower acceptance of the scientific consensus in every case. The conspiracist ideation scale included items with no direct connection to science, such as beliefs about Princess Diana's death and about the Apollo moon landings being staged. It also included what Lewandowsky and colleagues call "convenience conspiracies"—beliefs that specific scientific findings are themselves a hoax. The composite had a mean of two point thirty-seven on a five-point scale and strong internal consistency, with Cronbach's alpha of zero point eighty-four. The cognitive logic here is worth spelling out. Conspiracist thinking tolerates mutually contradictory explanations—different versions of a conspiracy can coexist even when they contradict each other. It treats isolated or anomalous evidence as more significant than the overall pattern. And it tends to be self-sealing: evidence that should falsify the conspiracy gets reinterpreted as proof that the cover-up is ongoing. Each of these features places conspiracist thinking in direct opposition to how science actually works, which depends on coherence, the weight of converging evidence, and the possibility of falsification. The genetically modified food result is particularly clarifying. Worldview did not predict genetically modified rejection at all, but conspiracist ideation did. That clean dissociation helps isolate the conspiracy effect from political content. It is not that genetically modified opponents happen to be conservative; it is that they are drawn to conspiratorial explanations regardless of where those explanations land politically. The bivariate associations are also striking. The climate hoax conspiracy item correlated at a correlation coefficient of negative zero point fifty-seven with acceptance of climate science—that's about a third of the variance explained by a single item. Conspiracy items tied to tobacco and HIV and AIDS correlated at negative zero point thirty-three and negative zero point eleven with acceptance of the relevant scientific propositions. These are not weak associations. Conspiracist thinking is doing a lot of work. All of this adds up to a picture where science rejection has at least two distinguishable roots. One is value-based and politically structured: people who endorse free markets resist science that implies regulation. This form of rejection is domain-specific—it shows up for climate science and not for genetically modified foods. The other root is epistemic and cross-cutting: a conspiracist cognitive style that treats scientific institutions as untrustworthy regardless of the topic. For science communicators, the split matters. When rejection is driven by worldview, targeting the regulatory framing can be helpful. Emphasizing scientific consensus is one documented lever—particularly for audiences already predisposed toward skepticism on political grounds. Since the rejection is grounded in values, it can, in principle, be engaged on its own terms. When rejection is driven by conspiracist ideation, direct correction is unlikely to be sufficient. Lewandowsky and colleagues are explicit: providing additional scientific information may amplify rejection rather than reduce it, because the new information gets absorbed into the existing conspiracy framework. The recommended alternatives are indirect—affirming the competence of scientific institutions, affirming audiences' own valued identities, and deploying broad simultaneous rebuttals that make conspiracist counter-explanations increasingly implausible. The deeper lesson is a methodological one for anyone thinking about this problem. Lumping all science skeptics together produces a distorted picture. A conservative who rejects climate science because they distrust environmental regulation is not the same kind of skeptic as someone who rejects genetically modified foods because they believe the agriculture industry is hiding the truth. The causes differ, the psychology differs, and therefore the interventions differ. Lewandowsky and colleagues show that you cannot treat science rejection as a single target without misdiagnosing it. A misdiagnosis, in communication as in medicine, leads to the wrong treatment. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.