Limits to inoculating against the risk of fake news: a replication study in Singapore during COVID-19

Abstract The COVID-19 pandemic laid bare the problem of fake news as one of the defining challenges of our time. The sudden proliferation of fake news and its direct impact on public health and safety led to increasing attention to pre-bunking interventions as a possible tool against the risks of fake news. These studies claimed that it is possible to use pre-emptive interventions such as games to induce cognitive resistance against the deception techniques deployed by fake new producers. We wanted to test if this method could be as effective in a non-Western context, and in an on-going catastrophic risk event. This paper presents the results of a replication study of Roozenbeek and van der Linden’s gaming experiment with certain modifications tailored to the case of Singapore in 2020 in the midst of the COVID-19 pandemic. We could not replicate the results of the original study. However, we found factors that could have accounted for the different results, including high levels of trust in English mainstream media and the government, and positive attitudes towards censorship. We also found that participants were most resistant against conspiratorial deception techniques but also more vulnerable to impersonation techniques. We reflect on what the results of our study say about the limitations of psychology-focused interventions and the need for a wider suite of interventions targeting different levels of analysis, including sociological factors and the risk context.