It's not "what % of college graduates get their news from this source", it's "what % of adults that get their news from this source have a college degree"

[–] Sergio@piefed.social 0 points 1 day ago (1 children)

OK you're right. the scales are:

on a scale of all US adults, the percentage who have a college degree
on a scale of all readers who primarily get their news from (given magazine), the percentage who have a college degree

So the scales are still different.

I'm guessing they'd make an argument that: "If the college graduate readership were distributed evenly across all news sources, then (given magazine) would have (the percentage of all US adults who have a college degree)." But the labels don't say that, which is why it is confusing.

[–] hesh@quokk.au 4 points 1 day ago* (last edited 1 day ago)

Imagine "All Americans" as one of the bars like the others - its just another cohort.