Who Falls for Fake News? Hungarian Study Points to Online Communities, Not Age Alone

Misinformation rarely spreads in a vacuum. It travels through networks of people, embedding itself in the daily habits, affiliations, and attention patterns of individual users. Yet most studies of who engages with false content online have relied on a single lens: either survey questions about beliefs and attitudes, or digital trace data showing what people actually click, like, and share. A new study of Hungarian internet users combines both, and its findings suggest that the online communities people join may matter more than nearly any personal characteristic in shaping their exposure to and engagement with misinformation.

The research, published in Humanities and Social Sciences Communications, was conducted by Júlia Számely and Elisa Omodei of the Department of Network and Data Science at Central European University in Vienna, together with Júlia Koltai of Eötvös Loránd University and the Centre for Social Sciences in Budapest. The team linked traditional survey responses with donated Facebook trace data drawn from a representative sample of the Hungarian internet user population. This dual approach allowed them to ask two complementary questions at once: who, in socio-demographic terms, is more likely to engage with misinformation, and what patterns of online behaviour accompany that engagement?

Hungary offers an unusually compelling setting for such an investigation. Over the past decade and a half, the country has drawn international attention for democratic backsliding, including the concentration of political power, pressures on independent journalism, and a media landscape in which pro-government outlets dominate much of the information environment. In societies where the media ecosystem itself is contested, understanding which individuals end up engaging with misinformation becomes not just an academic puzzle but a question with direct consequences for democratic deliberation. Misinformation in such contexts is not a fringe phenomenon; it can be woven into mainstream channels of political communication, making individual-level vulnerability harder to isolate and study.

Methodologically, the study stands out for its use of donated social media data. Participants in a representative survey agreed to share traces of their Facebook activity, giving researchers access to behavioural records that would otherwise remain invisible. Self-reported survey measures capture what people say about their media diets, their trust, and their beliefs, but such reports are notoriously prone to bias. Trace data, by contrast, reveal actual behaviour: which pages users follow, which interest clusters they belong to, and how they move through the platform’s ecosystem of communities. By joining the two data sources at the individual level, the researchers could compare the predictive power of conventional demographic variables against that of lived online behaviour.

The central finding is striking in its simplicity. Among the socio-demographic factors the study examined, only age showed a consistent association with misinformation engagement. Other conventional suspects, such as education level, gender, or income, did not emerge as reliable predictors once the full picture was considered. This challenges a widespread intuition, common both in public debate and in parts of the research literature, that misinformation is primarily a problem of particular demographic groups. The Hungarian data suggest that knowing a person’s place in the social structure tells us far less about their misinformation engagement than knowing where they spend their attention online.

What did matter, strongly and consistently, was membership in specific interest clusters on Facebook. Online platforms like Facebook organize social life into overlapping communities built around shared interests, identities, and information sources. The study found that these clusters act as powerful filters of information exposure: some ecosystems amplify engagement with misinformation, while others are associated with lower levels of such engagement. In other words, the communities a user belongs to shape which content crosses their screen, and repeated exposure within a like-minded environment can normalize engagement with dubious content. A user embedded in an information ecosystem saturated with unverified claims encounters, and therefore has far more opportunities to interact with, misleading material.

This framing recasts misinformation engagement less as an individual failure of judgment and more as a structural property of online environments. The algorithmic curation and social reinforcement that define modern platforms mean that users do not confront the information landscape alone; they see it through the lens of their communities. If those communities consistently circulate and endorse false or misleading content, engagement follows almost as a matter of course. Conversely, users whose communities anchor themselves in more reliable information sources appear buffered against such exposure. The practical implication is sobering: interventions aimed purely at individual characteristics, such as media literacy training targeted at particular age or education groups, may miss the mechanism that matters most, which is the informational character of the communities people inhabit.

The study’s broader methodological argument deserves equal attention. The authors emphasize that structural socio-demographic factors should be studied together with online behaviour to provide a more complete understanding of what drives misinformation engagement. Neither data source alone tells the full story. A survey can situate a person within their social context, but it cannot see the communities they actually inhabit online. Trace data can map those communities with precision, but without survey linkage they remain detached from the social characteristics of the people behind the accounts. Only by combining the two can researchers disentangle the relative contributions of social position and digital behaviour. The Hungarian study demonstrates that this combination can overturn assumptions built on either source in isolation, since the apparent demographic patterns may dissolve once community membership is taken into account.

The research also carries lessons for how misinformation should be studied in countries undergoing democratic strain. Hungary’s changing media landscape, marked by shifting ownership structures and politically aligned outlets, makes individual-level engagement with misinformation especially salient, the authors note. In such environments, the line between misinformation and partisan mainstream content can blur, and engagement with misleading content may reflect the structure of available information sources as much as any personal disposition. Studies conducted in more pluralistic media environments may not capture these dynamics, which makes evidence from Hungary valuable for understanding how information ecosystems shape behaviour wherever media independence is under pressure. The findings resonate with a growing body of work suggesting that the supply side of misinformation, meaning the networks and pages that produce and circulate it, deserves as much scrutiny as the demand side of individual susceptibility.

The study was conducted online between February and June 2023, approved by the Research Ethics Committee of the ELTE Centre for Social Science in Budapest, and carried out with explicit informed consent from all participants, who received detailed information about the objectives of the research and data management before taking part, and who retained the right to request deletion of their data at any time. The work was funded through the CEU Open Access Fund and the Hungarian Academy of Sciences Lendület Program. Its publication in an open-access journal ensures that the evidence, and the methodological template it offers, are available to researchers and policymakers grappling with misinformation far beyond Hungary’s borders.

The takeaway for anyone concerned about the health of the online information sphere is that community is destiny, at least where misinformation is concerned. Age may mark who is more or less likely to engage with false content, but the ecosystems people join on social platforms appear to do far more work in determining what they see and how they respond. Efforts to slow the spread of misinformation may therefore achieve more by focusing on the structure of online communities, the sources they amplify, and the pathways through which dubious content reaches them, rather than on demographic profiles alone. As social media platforms continue to sort users into increasingly self-reinforcing interest clusters, the Hungarian evidence serves as a reminder that the fight against misinformation is fought not just in individual minds, but in the architecture of the digital spaces where those minds gather.

Subject of Research: Socio-demographic and online behavioural drivers of misinformation engagement in Hungary using linked survey and Facebook trace data

Article Title: Socio-demographic and online behavioural drivers of misinformation engagement in Hungary: evidence from linked survey and social media data

Article References: Socio-demographic and online behavioural drivers of misinformation engagement in Hungary: evidence from linked survey and social media data. (n.d.). https://doi.org/10.1038/s41599-026-09002-0

Image Credits: AI Generated

DOI: 10.1038/s41599-026-09002-0

Keywords: misinformation, Hungary, Facebook, social media, survey data, trace data, online communities, interest clusters, demographic factors, information ecosystems, democratic backsliding, computational social science