Conspiracy Believers Are 91% Likely To Hold Incoherent Beliefs, Study Finds

Conspiracy Believers Often Hold Two Beliefs That Can’t Both Be True

Can someone genuinely believe Princess Diana was murdered by British intelligence and that she faked her own death? Logically, both can’t be true at once. Yet a sweeping new analysis of nearly 9,000 people suggests that for the most committed conspiracy believers, holding two mutually incompatible ideas at once isn’t a quirk, it’s practically the rule.

A research team from the University of Fribourg, the University of Bristol, and the University of Potsdam set out to settle a long-running debate: do conspiracy believers actually hold beliefs that clash with each other? Using a more rigorous math-based approach than earlier researchers had applied, they found that people at the extreme high end of a conspiracy-thinking scale had a 91% predicted probability of being logically inconsistent, versus just 3% at the low end.

That finding cuts against an influential 2023 study that concluded researchers had dramatically overestimated how often conspiracy believers hold incompatible ideas, putting the true figure below 3% of participants. This new paper, published in PNAS Nexus, argues that earlier study used a method that couldn’t detect individual-level inconsistency, and that a sharper tool brings the pattern back into focus.

Old Methods Missed Individual-Level Conspiracy Inconsistency

For years, researchers measured whether people who agreed more strongly with one conspiracy theory also tended to agree more strongly with a competing, incompatible one, treating a shared upward trend as evidence people held both beliefs at once.

That logic has a flaw. A positive relationship between two responses can show up even when almost nobody believes either claim. Someone who strongly doubts Diana was murdered and strongly doubts she faked her death will still produce responses moving in the same direction, creating a pattern with no genuine overlap in belief underneath it. Critics latched onto this flaw, reasonably so.

What critics missed, according to the new research, is that the fix isn’t abandoning the concept. It’s measuring incompatible belief at the level of individual people, rather than as an average trend across a group.

A Probability-Based Score Flags Inconsistency Person by Person

To fix this, researchers borrowed a framework from probability theory, converting each person’s agreement ratings into something closer to a probability estimate. Someone who rates a belief that Diana was murdered as very high, and a belief that she faked her death as also very high, ends up with two probability-like numbers that, added together, exceed what’s mathematically possible. Study authors compare it to trying to eat one and a half pizzas when only one pizza exists. They call this approach an incoherence score, and it flags a specific person as inconsistent instead of just tracking a trend across a group.

Researchers reanalyzed data from eight previously published studies, totaling 8,590 participants from Serbia, the United Kingdom, the Netherlands, and the United States. Each study had asked people to rate their agreement with pairs of competing conspiracy explanations for the same event, while also measuring how strongly they generally distrusted official accounts.

Applying that scoring method to the reanalyzed data, among participants who scored high on a conspiracy-thinking scale, 58% were inconsistent in at least one belief pair, compared to 11% among those who scored low. A computer simulation ruled out random or careless answering as the explanation; observed inconsistency was far higher than chance would produce, in the full sample and among high scorers alike.

Fictional Events Reveal Two Drivers of Conspiracy Inconsistency

A second study involved 469 U.S.-based, English-speaking participants recruited through an online research platform. Rather than using real-world conspiracy theories participants might already have opinions about, researchers invented fictional events set in Europe, stripping away the influence of prior knowledge or media exposure. Participants read short descriptions of ambiguous events, then rated the likelihood of two competing explanations, conspiratorial and otherwise, on a sliding scale from 0 to 100.

Results also turned up inconsistency outside conspiracy-themed explanations, suggesting part of the pattern may reflect a general tendency to miss logical constraints, rather than simply a motivated pull toward anti-establishment ideas.

Based on the combined results, the authors propose two distinct sources behind conspiracy-related inconsistency. One is content-specific: when both explanations match an anti-establishment worldview, believers are motivated to accept both, since each feeds a shared sense that the official story is wrong. The other is more general and, per the researchers’ interpretation, involves people high in conspiracy thinking being somewhat less likely to notice or act on logical constraints, even in neutral settings unrelated to conspiracy content.

Pointing Out Incompatible Beliefs Rarely Changes Minds

Past efforts to reduce incompatible conspiracy beliefs by pointing out the inconsistency directly haven’t worked especially well; people tend to rationalize incompatible ideas rather than abandon them. Study authors note one possible avenue: AI tools that engage people through personalized conversation could potentially help them notice when their own probability estimates don’t add up, though the paper stops short of claiming such tools have been proven effective at reducing conspiracy beliefs specifically.

Higher levels of conspiracy thinking were associated with a measurable departure from basic logical consistency, one that grew especially pronounced at the highest end of the scale. That raises real questions about how, or whether, incompatible beliefs can be dislodged once they’ve taken hold.

---

Disclaimer: This article summarizes findings from a peer-reviewed study and is intended for general informational purposes. It is not a diagnostic or clinical assessment of any individual’s beliefs or reasoning, and readers with concerns about their own thinking patterns or a loved one’s should consult a qualified mental health professional.

---

---