Misinformation alert. The outbreak is already dated to November, so October cases are not implausible *but* that paper has so many problems and so much hand-waving that it should not be cited, let alone made into news articles.https://twitter.com/borzou/status/1270250085084016641 …
-
Show this thread
-
zeynep tufekci Retweeted zeynep tufekci
I pointed out a few issues here. I don't mind the method (2003 SARS was identified partly due to online rumor tracking) but folks, hang on before making things into big news articles. Ask around. The paper is very short and will likely evolve.. a lot.https://twitter.com/zeynep/status/1270347472246919172 …
zeynep tufekci added,
zeynep tufekciVerified account @zeynepReplying to @chris_bailThis paper is not showing that at all, tbh. Diarrhea searches in August aren't proof for COVID. "(GI) symptoms are a unique feature of COVID19" is not accurate. Hospital parking lot increases do not indicate COVID. I read the paper and had to duck from all the hand-waving.2 replies 10 retweets 40 likesShow this thread -
zeynep tufekci Retweeted Bastian Purrer 🧢
Oh, wow. If true, this is nuts. I mean, the preprint paper is citing "Global Times" to support their assertion the market may not be the source of the virus. Internet searches and satellite data is fine to use, but there is no "top story" here. *headdesk*https://twitter.com/bastianpurrer/status/1270396180296597504 …
zeynep tufekci added,
5 replies 7 retweets 44 likesShow this thread -
We already have an earliest known case of mid-November. October cases are thus of course plausible. But there is no finding here. Nothing to report yet. *shouts into the void*
2 replies 9 retweets 56 likesShow this thread -
Look, reporting from preprints is fine, and there has been many good ones, but you really have to do the legwork and not rush, especially if the claim is very bold and consequential. Watch the caveats. Ask skeptical questions. Wait. Let papers evolve. *disappears into the void*
4 replies 9 retweets 57 likesShow this thread -
zeynep tufekci Retweeted Lena (Mimi) 🏠
This is not evidence at all! We have had antibody studies that show only a third of people who had genuine strong suspicion that they had COVID (their doctor thought so; they lived with COVID positive person) in *February* actually had COVID.https://twitter.com/athletespw/status/1270402000522489861 …
zeynep tufekci added,
Lena (Mimi) 🏠 @athletespwReplying to @zeynepAs I said, many athletes who were at the Military Games in Wuhan at end of October had symptoms similar to Covid19-symptoms. Here is an Italian article from a known paper https://www.corriere.it/sport/20_maggio_07/coronavirus-wuhan-quei-mondiali-militari-atleti-tornati-cina-quelle-strane-febbri-novembre-8924b39a-9037-11ea-b981-878bbbd902eb.shtml … To say "there were no Covid19-cases before mid-November" seems at least bold.3 replies 4 retweets 38 likesShow this thread -
zeynep tufekci Retweeted
Please also check this thread on the alleged early COVID wave. The curve fitting is reliant on maybe two days? Spike in parking in one hospital for two days? Could be anything. (I hope reviewers notice this). Meanwhile, very very thin story has gone viral. https://twitter.com/crushspread/status/1270570534581878784 …
zeynep tufekci added,
This Tweet is unavailable.1 reply 6 retweets 16 likesShow this thread -
Also note that if there was indeed a major wave of COVID in China in August 2019 (thin thin claim based just on searches for "diarrhea" but not anything respiratory but okay), there would be a lot of *export* of the disease, which we would be able to confirm via antibody testing.
1 reply 5 retweets 25 likesShow this thread -
zeynep tufekci Retweeted Will Ma
It's getting even shakier. Honestly, with all due respect to the authors (we're human, mistakes happen to everyone) and with my best and most generous interpretation of everything they claim, there was already nothing there in this paper without all this.https://twitter.com/WillMaStat/status/1270850256121540608 …
zeynep tufekci added,
Will Ma @WillMaStatI followed up with the Harvard team and found out “腹泻的症状” ("symptoms of diarrhea") was used, rather than “腹泻”("Diarrhea"), as the keyword in their search data analysis. Interestingly enough, the search volume of “腹泻” is actually decreasing during July-Dec 2019. pic.twitter.com/9jw6AcKisYShow this thread2 replies 4 retweets 21 likesShow this thread -
Replying to @zeynep
Google trends data is so hard to do right. The gap between what gets typed into a search bar and what's happening in the real world is massive. And it's so tempting to keep trying variations on your core search term until you get the signal you're looking for.
1 reply 1 retweet 5 likes
Especially in a language you don't speak.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.