A researcher puts ten letters in front of a participant. Each H stands for heads, each T for tails. There are runs and alternating pairs. The question is whether the sequence looks random or whether something appears to determine what comes next.
The researchers knew where the letters came from. They had generated a sequence of one hundred coin outcomes, constrained to include fifty of each, and broken it into sets of ten. Some participants saw more order than others. The same participants also answered questions about beliefs in conspiracies. Those who saw more pattern in the tosses tended to endorse more of the conspiracy claims.
A row of letters is a long way from a real scandal. Nobody was hiding a document or changing an account under oath. That simplicity is what makes the experiment worth looking at. It gives an investigator something rarely available in public life: independent knowledge of how the apparent pattern was produced.
Where suspicion begins
Imagine finding the same name in two government files. It is reasonable to ask what connects them. Perhaps the person worked on both matters. Perhaps the same official processed thousands of documents. The difference lies in evidence outside those two appearances. Once the coincidence becomes the explanation, it becomes harder to ask how often the name ought to turn up by chance.
A conspiracy theory proposes that people secretly coordinated to bring about an event. People sometimes do coordinate in secret. That makes the question of how to test any particular claim more important. A psychological study can examine why a person finds an explanation persuasive. Its results cannot decide what actually happened in the event under discussion.
What an explanation can offer
Psychologists Karen Douglas, Robbie Sutton and Aleksandra Cichocka reviewed research on three broad motives. Someone may want an explanation for a confusing event. They may want to feel safe or able to influence what happens. Or they may want to make sense of how a group they belong to is treated by others.
Those are different experiences. A reader trying to reconcile two dates in a report and a reader convinced that an institution has dismissed people like them may arrive at the same website by entirely different routes. Even when a secret-plot explanation appeals to a particular need, there is a further question. Does adopting it actually bring a greater sense of understanding, security or belonging? The review cautions that attraction to an explanation is easier to demonstrate than lasting satisfaction.
Testing the scale of the association
A later review combined 279 independent studies involving more than 137,000 participants. It found associations across all three groups of motives. Their average size was modest, and results varied considerably across studies. That matters when moving from a research paper to a conversation with one individual.
An average describes a relationship across many observations. It cannot tell you why the person across the table distrusts one investigation or which particular fact might change their view. Nor can a correlation tell us on its own which came first. Feeling uncertain might increase interest in an explanation, while a frightening explanation could also leave someone feeling less secure.
What happens when someone offers contrary evidence?
One familiar story about conspiracy beliefs says corrections simply make people dig in. That prediction can be tested. In a 2025 experiment involving 955 participants, each person reported a conspiracy belief or another belief lacking adequate evidence. A language model then discussed the specific claim with them. Researchers varied whether participants were told that the speaker was an AI tool or a human expert. They also varied the tone of the conversation.
Participants reported lower confidence in their beliefs after the conversation. The researchers found no significant difference in that change between the messenger descriptions or the tested tones. The study compares different versions of a particular intervention. It cannot promise the same result from posting a link or arguing with a relative.
What distinguished the conversation was its attention to particulars. A person convinced by an unusual photograph has a question about the image’s origin and context. Someone persuaded by a timeline needs a different kind of answer. A blanket claim about people who believe conspiracies risks ignoring the very evidence they think they possess.
Frequently asked questions
Why do people believe conspiracy theories?
Research identifies several pathways, including a search for explanations, concerns about safety or control, and social identification. Associations differ among studies and individuals.
Can you recognize a believer from intelligence or education?
These studies provide no individual diagnostic test. Someone’s qualifications cannot settle whether a specific allegation is supported.
Does correcting a claim always backfire?
No. In the controlled conversation study the average participant’s stated belief declined. That finding leaves open which method works best in other settings.
Do real cover-ups change how a new claim should be evaluated?
Past misconduct can give someone a concrete reason to ask for documentation. Establishing a new allegation still requires evidence about the people, decisions and events it describes.
Why can a story circulate so easily online?
One person can pass on an arresting detail without carrying along its original source or competing explanations. To understand its spread on a particular platform would require data about what people actually saw and shared. The coin-toss study makes a smaller, more revealing point: seeing a pattern can be the beginning of an investigation, but the pattern cannot testify to who put it there.
Research Sources
The following resources were consulted to verify and update this article.
- The psychology of conspiracy theories — Reviews the roles of uncertainty, security and social identity, and the limits of claims about fulfilled needs.
- Reasons to believe: A systematic review and meta-analytic synthesis of the motives associated with conspiracy beliefs — Combines 279 independent studies and measures modest, variable associations across three classes of motives.
- Connecting the dots: Illusory pattern perception predicts belief in conspiracies and the supernatural — Reports the coin-sequence task and the relationship between perceived patterns and conspiracy beliefs.
- Boissin and colleagues, Dialogues with large language models reduce conspiracy beliefs even when the AI is perceived as human — Reports the 2025 preregistered experiment comparing messenger descriptions and conversational tone.
🔎 If this story stayed with you, the author suggests these real cases next:Explore more Conspiracies & Theories stories here: View all Conspiracies & Theories stories →
- The proven secret programs that taught the public to doubt official stories
- The CIA mind-control program that turned suspicion into documented history
- The white streaks overhead and the theory that keeps reinventing itself
- The hidden world beneath our feet that some still insist is real
- The assassination case that taught generations to distrust the official version
Related reading
Watch This Story on YouTube
Prefer to watch? See the full story on our YouTube channel.
