They used Amazon’s Mechanical Turk to categorize the “slant” of various tweets selected by a computer program. The tweets were labeled as slanted “D” (democrat)

They used Amazon’s Mechanical Turk to categorize the “slant” of various tweets selected by a computer program. The tweets were labeled as slanted “D” (democrat), “R” (republican), or “N” (neutral). Looking at the tweets[1] (table S2 on page 6, fair warning: PDF) I think the use of mechanical turk really undermines the research they’re trying to do. A fair number of these tags are just plain wrong. For example, how are these “neutral”? > Most of the money that Bernie Sanders raised immediately after announcing his presidential bid was donated in the Venezuelan currency, leading his campaign to dramatically overstate the size of the contribution. > A SWAT team was sent to arrest a man after Facebook reported him for holding extremely conservative views. [1]https://advances.sciencemag.org/content/advances/suppl/2021/05/28/7.23.eabf1234.DC1/abf1234_SM.pdf https://advances.sciencemag.org/content/advances/suppl/2021/...