3 ms·
> Among other findings, the researchers observed that the accuracy of news sources the Twitter users retweeted promptly declined by roughly 1 percent in the nex
by jsight 5y ago
> Among other findings, the researchers observed that the accuracy of news sources the Twitter users retweeted promptly declined by roughly 1 percent in the next 24 hours after being corrected.
1 percent? How do you measure a 1 percent decline in news source quality? This sentence throws doubt at the whole idea, IMO.
- nicklecompte 5y agoThe paper is online: https://dl.acm.org/doi/pdf/10.1145/3411764.3445642 https://dl.acm.org/doi/pdf/10.1145/3411764.3445642 > To measure users’ subsequent behavior after receiving the correc- tion, we focused on three main outcome variables. Most importantly, we considered the quality of news content shared by the users. We quantifed the quality of news content at the source level using trust- worthiness scores of news domains shared by the users based on a list of 60 news domains rated by professional fact-checkers (this list contains 20 fake news, 20 hyperpartisan, and 20 mainstream news outlets where each domain has a quality score between 0 and 1) [39]. A link-containing (re)tweet’s quality score was defned as the quality of the domain that was linked to. (Quality scores could not be assigned to tweets without links to any of the 60 sites.) Citation [39] is Pennycook,G. & Rand,D.G. “Fighting misinformation on social media using crowdsourced judgments of news source quality.” I am not sure what “crowdsourced” means here because I didn’t read all the citations. But it sounds like they have a plausible metric. It’s always very dumb to criticize science based on a press release.
- jsight 5y agoI'm not sure what you are arguing here. Do you think 1% is a significant finding? What was the margin of error?
- nicklecompte 5y agoI don’t really know if it’s a significant finding. My point was that you asked “how you measure a 1 percent decline in news source quality?” and suggested that your personal confusion about such a thing throws the entire paper into question. Instead of having such strong judgments, why not just read the paper? I am personally not sure how astronomers measure gravitational redshift with such apparent accuracy. But my ignorance does not mean astronomers are a bunch of frauds. It means I haven’t done the required reading.
- jsight 5y agoDoubt is not a strong judgement. I was hoping someone had the details and would clarify. I appreciate that you did that as you've turned my doubt into certainty. They did not precisely measure a 1% change in news source quality.
- ChrisLomont 5y agoThis thread epitomizes the paper.
- mistermann 5y agoGravitational redshift didn't convince you?