Health Nerd Profile picture
17 Mar, 9 tweets, 2 min read
Something I think about a lot is that studies don't get retracted because they're bad, they get retracted because they are famous
Don't get me wrong, they are ALSO bad. It takes a truly awful study to get a scientific journal to wrest itself free of apathy and inertia to take some action

But there are 1,000s of woeful papers
Thing is, no one is paid to catch bad research. It is a thankless, time-consuming task that at best earns you the mistrust of most of your peers

And so, no one does it methodically
I don't have a ton of free time, and I don't go out searching for every error-filled study. I suspect most other people involved in this area are the same based on various conversations
For someone to notice a bad paper, it has to be spread around quite a bit. In the news. Causing a social media storm

Most bad research isn't popular like that. It never gets read by more than a handful of people
Moreover, if the study is not getting news, NO ONE CARES

Journals aren't quick to act when the study is literally impacting policy in a way that could kill thousands - imagine how slow they move when it's a nothing of a study that's been read twice in 10 years
And that's why I say only famous research gets retracted. There are exceptions to the rule, of course, but generally it holds surprisingly well
Anyway, when people get mad about how the papers that are retracted tend to be the big flashy studies I just think that they are ignorant about how the whole error-checking in science thing works
Also, I cannot stress enough how foolish it makes you look if you say "why are you only going after THESE studies and not all the bad ones?"

Error detection in science is a job that pays in sleepless nights and ennui. No one pays us to do any of this

• • •

Missing some Tweet in this thread? You can try to force a refresh

Keep Current with Health Nerd

Health Nerd Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!


Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @GidMK

18 Mar
Very interesting study out of Denmark looking at SARS-CoV-2 reinfections:

- 0.65% symptomatic reinfections after 7 months
- in sensitivity analysis this doubled to 1.2%
- estimated ~80% short-term protection against reinfection Image
Studies like this make me very jealous of my Nordic colleagues. The authors had access to linked data for *the entire country of Denmark*, which is a pretty enormous strength of the research Image
Basically, they looked at every PCR test done in the first wave, and followed up every person to see if they had tested positive in the first, second, or both waves

Of those in the sample, 0.65% were infected twice
Read 7 tweets
16 Mar
You may have seen the massive viral tweets about how staying at home doesn't prevent COVID-19 deaths

These were based on a paper with what we think are quite significant flaws

Our full critique now preprinted here:
The basic explanation here is that the original article looked at whether Google "residential" mobility data was correlated with COVID-19 death rates, and found no association
There are significant drawbacks with that methodology, some of which I outlined in a thread

Read 8 tweets
15 Mar
Your daily reminder that "I'm pro-vaccine except for THIS one" is literally the most common anti-vax line there is
The second most common line is "I'm pro-vaccine but I'm also pro informed choice" usually followed by a slew of lies and misinformation portraying vaccines as dangerous
So many replies missing the point. There's a big difference between common talking points and actions - most anti-vaccine advocates SAY they are only against one vaccine but then come up with similar arguments against ALL of them
Read 4 tweets
14 Mar
People ask this question a lot, and I think it's actually worth an answer, so my thoughts:

Have lockdowns caused large numbers of excess deaths? 1/10
2/10 The MAJOR caveat here is that we are only looking at short-term impact. Long-term impact is going to take a while to assess
3/10 What would we expect to see if lockdowns caused large numbers of deaths?

Well, lots of COVID+lockdown would = many deaths, and no COVID+lockdown would ALSO = many deaths
Read 11 tweets
13 Mar
According to Prof John Ioannidis, the cause of 2.6 million* COVID-19 deaths was...pretty much everything except the disease

*likely a substantial undercount
He then clarifies that these were not the only cause of death, but of course without these issues COVID-19 would've been a nothing of a problem
I had hoped that this was fake, because this argument is just shockingly bad, but nope this is a real slide the professor actually presented
Read 5 tweets
11 Mar
Interesting update on this paper published that purported to show that staying at home doesn't reduce COVID-19 deaths: less than a week after publication it already has a warning from the editors
Also, the authors appear to have responded to my twitter thread that was automatically uploaded to Pubpeer, which is pretty fantastic. Not sure this helps their case tho
"This is the best data available" is not really a defense about using inadequate data. If you don't have the data to answer a question, then it's not a surprise that your study fails to find an effect I think
Read 4 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!