What I found more interesting is that "procedural news knowledge" made participants less likely to say that they would share a headline if that headline was false. But maybe it just helps with identifying proxy signals like which outlet the headline was most likely sourced from.
