Causal Inference in Misinformation Research: A Proposed Roadmap for Studying Behavioural Effects

Concerns about the societal consequences of misinformation have intensified in recent years, particularly regarding its potential impact on democratic processes, public health, and collective decision-making. Despite these concerns, the central question of when and how misinformation exposure might directly affect behaviour remains surprisingly underexplored. Drawing on a real-life case study, we show how pre-existing individual differences may simultaneously drive both engagement with misinformation and the behaviours it is assumed to cause, rendering causal conclusions misleading. We situate this problem within a wider empirical landscape in which field, quasi-experimental, and laboratory studies each provide partial but incomplete evidence regarding behavioural effects. Building on this critique, we provide a roadmap for future research, specifying methodological and conceptual requirements for establishing credible causal claims, including improved alignment between research questions and design and greater use of behavioural outcomes. We argue that building a rigorous body of causal evidence is an essential step towards more effective interventions.