Fighting misinformation among the most vulnerable users.

Misinformation undermines trust in the integrity of democratic elections, the safety of vaccines, and the authenticity of footage from war zones. Social scientists have proposed many solutions to reduce individuals' demand for fake news, but it is unclear how to evaluate them. Efficacy can mean that an intervention increases discernment (the ability to distinguish true from false content), works over a delay, scales up, and engages users. I argue that experts should also consider differences in exposure prevalence before declaring success. Misleading content makes up a small fraction of the average person's news diet, but some groups are at increased risk - conservatives and older adults see and share the most fake news. Targeting the whole population (universal prevention) could concentrate benefits among the users who already see the least misinformation to begin with. In complement to these approaches, we should design interventions for the people who need them most - conservatives and older adults (selective prevention), as well as users who have already shared low-quality news (indicated prevention).