On college majors and voting preferences
(media.communities.win)
You're viewing a single comment thread. View all comments, or full comment thread.
Comments (33)
sorted by:
The assumption is that the remainder follow a roughly similar split - they're unlabeled, so assuming the same valid-fake distribution as the rest of the dataset is the best we can do.
Presumably. The proportions aren't listed, though, and they don't map cleanly to the other dataset.
That's not right. The 2016 election, in which graduates (docs excluded, since we're looking at bach degree stats, and they're negligible in number regardless) skewed more left-wing than usual, was still split 49-45, with 'some college' splitting 43-53 in the other direction.