Inoculation against misinformation in Australia: a replication study

Misinformation has the potential to damage the public’s trust in governments’ policies and communications. In this large three-arm parallel online randomised controlled trial, we aimed to test the efficacy of two commonly applied misinformation interventions: false tags and psychological inoculation.Australian residents (n = 1,815) were randomly allocated to one of three experimental arms: control, false tags or inoculation. We exposed participants to 30 real-life social media posts, 15 of which were misinformation. While previous misinformation experiments have predominantly used perception- or intention-based outcome measures, our social media platform simulation measured engagement with misinformation, a more ecologically valid approach. Specifically, the ‘reactions’ of participants to each post were recorded. Using pre-registered mixed-effects logistic regression models, we found that inoculation conferred the most protection against misinformation, significantly reducing ‘likes’/‘loves’ and ‘reactions’ compared to the control. Inoculation also suppressed ‘likes’/’loves’ significantly more than false tags, the approach currently used by Meta. The results of this experiment suggest that inoculation has the potential to attenuate engagement with misinformation in Australia. The results of this study may be used by Australian regulators and social media platforms to inform future policy decisions.