“…Participants rated the real news headlines as substantially more accurate ( M = 2.64) than the fake news headlines ( M = 1.95), t (401) = 20.9, p < 0.001, d = 1.04—that is, participants were fairly good at discerning fake from real. Replicating the findings of Pennycook and Rand (), headline type (0 = real, 1 = fake) interacted with CRT score (continuous), β = −0.297, t (400) = −4.23, p < 0.001, such that CRT performance was negatively correlated with lower perceptions of fake news (but not real news) accuracy (Table ; descriptive statistics can be found in SM). In contrast, removing the source from the news stories had no effect on perceptions of accuracy (Source: M fake = 1.98, M real = 2.65; No Source: M fake = 1.93, M real = 2.62), regardless of CRT (i.e., no significant main effect or interactions, p > 0.05 for all).…”