Rich-poor gap in spotting misinformation narrows among teens who read well

The study tests an interaction. Jabbari and Hossain asked whether the payoff from reading depends on family background. In their data it does: the link between reading achievement and misinformation resistance was strongest among disadvantaged students, and stronger reading gave students with fewer resources a bigger advantage ^[2]. "Reading skills could play a compensatory role here," Jabbari said ^[4].

The sample is broad. Spread over 49 countries, 232,901 students works out to an average of about 4,750 per country ^[18]. A sample that size makes the pattern precise, but every student was measured once, in the 2022 round ^[12], so reading ability was observed and never assigned.

The thing this doesn't tell you is whether teaching a poor student to read better would make that student harder to fool. Strong readers from poor families may differ from weak readers in other ways, such as their schools or their teachers, and one snapshot cannot separate those factors from reading. Jabbari's own explanation lists the kind of background factors that go with income. "Students from more advantaged families often have other supports, like parents who discuss the news, access to credible sources and norms that encourage skepticism," he said ^[5].

His argument is that reading makes up for those supports where they are missing. "As a result, the ability to comprehend, compare and evaluate texts may carry more weight for low-income students," he said ^[7]. He also said teenagers' online behavior depends on their social environment as well as on their ability to evaluate information ^[11]. That second point cuts both ways. If the environment matters, the strong readers in poor families may also be the ones in better environments.

Effect size is harder to pin down. The press release does not report how large the gap is at either end of the reading scale. Jabbari's word for the narrowing is "considerably" ^[3]. The outcome also rests on what students reported about their own habits ^[16], ^[17]. A self-reported habit is a useful signal. It is still a different quantity from how often a teenager actually forwards a fabricated post.

I think the evidence is strong enough to justify the experiment it cannot replace. That would be a trial that assigns extra reading instruction to schools serving low-income students and then measures misinformation resistance. Jabbari argued that 15-year-olds are the right age to reach, because they are increasingly handling online information on their own while still in school ^[10]. "This means we can still reach nearly all of them through classroom instruction before they enter adulthood and begin major milestones, like voting," he said ^[10].