How the Label ‘Conspiracy Theory’ Shapes What Groups Decide to Believe

When a claim sounds outlandish, calling it a conspiracy theory can end the conversation before it starts. New research from Flinders University in Australia suggests that this everyday move is more than a dismissive habit: the label actively shapes how groups reason together, and it can determine whether people engage seriously with contested information or simply laugh it out of the room. The study, published in Frontiers in Social Psychology, offers one of the most direct looks yet at how conspiracy beliefs are negotiated in real time, in the presence of other people, rather than formed silently inside individual minds.

Psychologists Cassandra Quarisa and Emma Thomas set out to address a gap in a literature that has grown enormously over the past decade. Hundreds of studies have catalogued the personality traits, motivations, and cognitive styles associated with believing conspiracy theories, but most of that work treats belief as a private possession, measured by asking individuals to rate statements on a scale. Yet conspiracy theories are inherently social artifacts. They spread through conversation, they are shared for impression management or social connection, and they are often transmitted faster online than factual content. What has remained poorly understood is the moment of deliberation itself: what people actually say when they decide whether an alternative explanation of events deserves belief.

To capture that moment, the researchers ran two experiments with United States residents recruited through Amazon Mechanical Turk. In the first study, conducted in August 2021 during the initial vaccine rollout, 244 participants read balanced excerpts presenting both the publicly accepted account of the origins of COVID-19 and a conspiratorial account claiming that wealthy elites such as Bill Gates and Dr. Anthony Fauci had engineered the pandemic for profit. In the second study, conducted in February 2022 amid booster campaigns, 374 participants read paired accounts concerning vaccination, including the claim that pharmaceutical companies were faking data on vaccine safety and efficacy. After reading, participants were randomly assigned to one of two conditions. Half deliberated alone, writing a short reflection paragraph within seven minutes. The other half entered live, unmoderated text-based group chats with two or three other participants and had ten minutes to reach a joint decision about whether they agreed with the conspiratorial account.

The resulting corpus, roughly 127 words per individual reflection and around 400 to 460 cumulative words per group chat, was analyzed using Reflexive Thematic Analysis, a qualitative method that codes textual data against a theoretical framework while remaining open to emergent patterns. The framework drew on a recent causal definition of conspiracy theories, which holds that such beliefs share five structural properties: they are oppositional to publicly accepted knowledge, they describe malevolent or forbidden acts, they ascribe agency to powerful actors, they are epistemically risky in that they lack evidentiary trails, and they are social constructs capable of creating social realities. The researchers adopted a critical realist epistemology, treating the written deliberations as empirical observations from which underlying mechanisms of belief formation could be inferred.

Five major themes emerged across the two studies. Three appeared consistently regardless of topic: plausibility, trust, and harm. Plausibility captured participants’ assessments of whether the claims lacked direct evidence, contradicted accepted knowledge, or simply sounded like a conspiracy theory. Trust concerned perceptions of the alleged conspirators, whether they were seen as capable and willing to orchestrate a plot, and whether institutions such as science, government, and pharmaceutical companies deserved confidence. Harm centered on whether the conspirators’ actions would damage the public or whether their motivations might be benign. Two additional themes surfaced only in the vaccination study: uncertainty, reflecting the sense that information about the new COVID-19 vaccines was insufficient, and autonomy, reflecting the feeling that mandates violated personal choice.

Crucially, the same structural features were recognized by believers and skeptics alike; what differed was how those features were construed. Participants who rejected the conspiracies typically negated one or more elements explicitly. In one group chat, members acknowledged that pharmaceutical companies profit from vaccines but quickly established a norm that regulation and transparency mitigate the risk, concluding that profit motive is not evidence of a nefarious plot. Others argued that the scale of coordination required, involving scientists and governments across every major country maintaining secrecy for years, made the theories implausible. Participants who endorsed the conspiracies, by contrast, treated the same features as accurate descriptions of the world, often drawing on personal experiences of censorship or perceived suppression to reinforce the idea that hidden enemies were at work. Notably, when the alleged conspirator was an institution rather than an individual, participants tended to propose structural solutions such as oversight rather than assigning individual blame.

The most striking finding concerned the discursive power of the label itself. In the pandemic study, participants in group chats rapidly converged on the judgment that the alternative account was a conspiracy theory, and this labeling had cascading effects. Once a claim was tagged as conspiratorial, groups developed norms of ridicule, with members making satirical and disparaging remarks about conspiracy theorists as a category of irrational others. Dissent was swiftly suppressed. In one exchange, a participant who attempted mild devil’s advocacy about a documentary linking a public figure to vaccines was quickly shut down by others declaring it just hype from conspiracy theorists. The researchers observed that this pattern fostered an anti-conspiracy identity, in which distancing oneself from anything remotely conspiratorial became a marker of group belonging, and the label functioned as a proxy for truth or falsity rather than prompting critical evaluation of the claims themselves.

The vaccination study told a different story. There, participants were far less likely to apply the conspiracy label, partly because elements of the claims, such as the existence of adverse effects or the profit motive of pharmaceutical companies, seemed observationally true. Group chats in this study displayed markedly more openness, with members affirming one another’s perspectives and adopting nuanced positions that accepted some aspects of the claims while rejecting others. Themes of uncertainty and autonomy emerged almost exclusively in these group discussions, and groups that could not reach a unanimous decision were characterized by norms that permitted expressions of doubt without treating them as endorsement of the conspiracy. Participants balanced their desire for certainty with tolerance for ambiguity, often framing vaccination as a matter of personal choice rather than adopting conspiratorial reasoning.

The authors caution that their findings come with limitations. Fewer than 14 percent of participants endorsed the conspiracies, so the analysis disproportionately represents skeptics, an imbalance that mirrors real-world opinion distributions but constrains conclusions about how believers negotiate their views socially. The unmoderated chat format meant the researchers could not control group composition, and the politically salient figures named in the stimuli may have shaped responses. Still, the researchers argue that artificially normalizing such skewed distributions would misrepresent the phenomenon, noting that contemporaneous polling found between 10 and 25 percent of Americans believed claims about Bill Gates and the pandemic.

The broader implication is that group interaction is neither inherently a cure for nor a cause of conspiracy belief. Instead, the norms that emerge within a deliberating group, whether norms of ridicule and consensus or norms of open dialogue and critical scrutiny, determine the outcome. When the conspiracy label is deployed pejoratively, it can shut down nuanced engagement and even produce backlash, with people perceiving censorship as confirmation that a plot is being hidden. When groups instead create space for transparent discussion, healthy skepticism, and independent critical thought, participants take a more active role in their decision-making. The researchers suggest that such open, nuanced dialogue may mitigate the backlash effects of the pejorative conspiracy label while empowering people to weigh contested claims on their own merits, a finding that resonates with recent work showing that even brief, non-confrontational dialogues with AI chatbots can durably reduce conspiracy beliefs by encouraging critical thinking rather than ridicule.

Subject of Research: How conspiracy attributions function as discursive tools in individual versus collective sense-making about COVID-19 and vaccination

Article Title: “Seems like the conspiracy theorists are at it again”: conspiracy attributions as a discursive tool in collective (vs. individual) sense-making

Article References: Quarisa, C. M., & Thomas, E. F. (2026). “Seems like the conspiracy theorists are at it again”: conspiracy attributions as a discursive tool in collective (vs. individual) sense-making. Frontiers in Social Psychology, 4, Article 1768090. https://doi.org/10.3389/frsps.2026.1768090

Image Credits: AI Generated

DOI: 10.3389/frsps.2026.1768090

Keywords: conspiracy theories, collective sense-making, social psychology, COVID-19, vaccination, group deliberation, discursive tools, thematic analysis, misinformation, social norms, trust in institutions, qualitative research