Article in Chronicle of Higher Education on retraction of Covid vax paper

[Dismal]

This was from a month ago, but I keep seeing social media references to it. The author, an economist at Michigan State, has been interviewed multiple times expressing bafflement that the journal accepted then retracted his paper.

https://www.chronicle.com/article/this-questionable-study-caught-fire-in-anti-vaccine-circles-how-did-it-get-through-peer-review?sra=true&cid=gen_sign_in

People are surveyed and asked whether they know anyone who died from the Covid vaccine. A surprising number said yes. Then this rumor data are compared to government stats and the author concludes that the official statistics should be investigated due to the discrepancy.

[Caracal]

I can't read the article because it's behind a paywall, but it does seem bizarre that this wouldn't just be rejected out of hand, at least if the author is claiming that this is a useful way to look at vaccine safety.

I can't imagine there's any real way you could extract useful data on prevalence of anything by asking people about whether something has occurred to other people, outside of their immediate family. Obviously, what we hear is dependent on information networks and those networks aren't just neutral tubes but filtered through all kinds of social factors. If you ask me "are you aware of anyone you know who has declared bankruptcy" I would say no. If you asked me if anyone I had personally known had died of a drug overdose I would say yes. You couldn't do this survey, however and use it to estimate the prevalence of drug overdoses and bankruptcies. My next door neighbor when I was a kid died of a drug overdose and I learned about this on social media, but it's hard to imagine I would have learned about it if had declared bankruptcy.

That's before you even get into the particular issues with Covid vaccines where there's widespread disinformation that might encourage people to believe that someone died of a vaccine when it had nothing to do with their death.

[secundem_artem]

The journal in question allows readers to read the peer reviews.  They were utterly laughable.  This was not a well done study and should never have made it through the review process had the reviewers and editor been even slightly competent.

[Hegemony]

They should do a study on the prevalence of kidneys removed leaving the victim in the motel room bathtub in a pile of ice. (Widespread urban legend.) I'm sure the number of people who have heard of someone that's happened to differs from the number of times it's been officially reported. Needs looking into! A cover-up!

[apl68]

Wasn't a catastrophic failure of peer review at a noted journal responsible for that article that started the whole autism-linked-to-childhood-vaccines business some years back?

I guess a lot of people are just determined to help open the First Seal and unleash the White Horseman on the world once again.  As if the late pandemic wasn't already enough.

[Mobius]

Can you name someone you know who died due to a Covid vaccine would get some interesting responses. I'm sure the Buffalo Bills player who collapsed on the field would get a few mentions (some people believe he was replaced by a body double).

[Hibush]

Can you name someone you know who died due to a Covid vaccine would get some interesting responses. I'm sure the Buffalo Bills player who collapsed on the field would get a few mentions (some people believe he was replaced by a body double).

There is a place for this kind of thing in behavioral economics journals because false perceptions influence economic decisions. How would a paper addressing this particular question fare at a non-predatory journal with open review?

[jerseyjay]

There is much about the whole episode I found bizarre (an economist publishing in a medical journal, the superficial "peer reviews," the superficial article itself).

If I understand the article correctly, the author did a survey that found that people who know somebody who died of Covid are more likely to get vaccinated than people who know (or think they know) somebody who became ill because of the Covid vaccine. That seems plausible, although I am not sure what it proves. The insights, as it were, of the paper would seem to be on public perception and attitudes rather than the safety or efficacy of the vaccination.

But then the author points out that there is a large gap between people who say they know somebody injured by the vaccine and the actual (reported) numbers of people injured. This is probably true, too. It is also true for UFO abductions, satanic possessions, and reincarnation. On a logical level, it is possible that there is under-reporting (either by mistaken reporting, conspiracy, etc). It is also possible that people don't know what they think they know. Again, this might be interesting from the perspective of perception and attitudes, but not at all useful in trying to figure out how many were injured from the vaccine.

On a less devious plane, assuming that there is no conspiracy and people are not just ignorant, there is no concrete relationship between knowing somebody who falls into a certain group and the size of the group: the number of faculty members at my school who know somebody with a Harvard PhD is much greater than the number of people with Harvard PhDs. Or, we all know the university president, but there is still only one university president. 

But all this obvious. Which why the whole episode is bizarre.

[Puget]

On a less devious plane, assuming that there is no conspiracy and people are not just ignorant, there is no concrete relationship between knowing somebody who falls into a certain group and the size of the group: the number of faculty members at my school who know somebody with a Harvard PhD is much greater than the number of people with Harvard PhDs. Or, we all know the university president, but there is still only one university president. 

But all this obvious. Which why the whole episode is bizarre.

Yeah, this is the weirdest part-- even if people were reporting accurately (which they probably weren't, but even if they were), as you note you obviously can't compare "know someone who" to "am someone who" numbers. I thought economists were supposed to be good at math and logic?  /snark

[dismalist]

There is much about the whole episode I found bizarre (an economist publishing in a medical journal, the superficial "peer reviews," the superficial article itself).

If I understand the article correctly, the author did a survey that found that people who know somebody who died of Covid are more likely to get vaccinated than people who know (or think they know) somebody who became ill because of the Covid vaccine. That seems plausible, although I am not sure what it proves. The insights, as it were, of the paper would seem to be on public perception and attitudes rather than the safety or efficacy of the vaccination.

But then the author points out that there is a large gap between people who say they know somebody injured by the vaccine and the actual (reported) numbers of people injured. This is probably true, too. It is also true for UFO abductions, satanic possessions, and reincarnation. On a logical level, it is possible that there is under-reporting (either by mistaken reporting, conspiracy, etc). It is also possible that people don't know what they think they know. Again, this might be interesting from the perspective of perception and attitudes, but not at all useful in trying to figure out how many were injured from the vaccine.

On a less devious plane, assuming that there is no conspiracy and people are not just ignorant, there is no concrete relationship between knowing somebody who falls into a certain group and the size of the group: the number of faculty members at my school who know somebody with a Harvard PhD is much greater than the number of people with Harvard PhDs. Or, we all know the university president, but there is still only one university president. 

But all this obvious. Which why the whole episode is bizarre.

On a less devious plane, assuming that there is no conspiracy and people are not just ignorant, there is no concrete relationship between knowing somebody who falls into a certain group and the size of the group: the number of faculty members at my school who know somebody with a Harvard PhD is much greater than the number of people with Harvard PhDs. Or, we all know the university president, but there is still only one university president. 

But all this obvious. Which why the whole episode is bizarre.

Yeah, this is the weirdest part-- even if people were reporting accurately (which they probably weren't, but even if they were), as you note you obviously can't compare "know someone who" to "am someone who" numbers. I thought economists were supposed to be good at math and logic?  /snark

I read the paper, and some criticism of it. JJ is correct. The substantive part seems fine to me, but not all that much to write home about: People who claim to know of deaths from vaccination get vaccinated less. Coulda' been discussed.

The author goes a bridge too far, a bloody big one, when he uses his data on who personally knows  vaccination deaths to compare it to official data. He finds a large discrepancy and says somebody should investigate. This is presented like an afterthought would be.

The paper got picked up by anti-vaxxers and that caused the brouhaha. A critic noted that  the author himself was an anti-vaxxer.

Competent editors would have had the afterthought excised.

As I've often claimed, the sociology of science is much more interesting than the philosophy of science!

[Caracal]

I read the paper, and some criticism of it. JJ is correct. The substantive part seems fine to me, but not all that much to write home about: People who claim to know of deaths from vaccination get vaccinated less. Coulda' been discussed.

Sure, if it's acknowledged that this is probably about the way people interpret evidence to line up with their own beliefs. To be fair, I certainly do the same thing. Eric Clapton claims that he developed a severe and debilitating illness as a result of getting the vaccine. I don't believe him. Why? Well, I could claim that my judgement is based on the very low statistical likelihood of having serious adverse reactions to the vaccines. But, that doesn't quite work. If Clapton said he had been bit by an alligator, I'd believe him, all things being equal, even though very few people get bit by alligators.

What's really happening here is that I think Clapton has a series of beliefs about the vaccines and has constructed a narrative that blames them for whatever health problems he's dealing with. My conclusion isn't really based on a particularly careful scrutiny of data. Very rarely, some people do have serious adverse reactions. Some of those people might develop beliefs about the vaccine based on their own experiences rather it working the other way around.

Of course, if we are trying to figure out what the actual prevalence of these reactions is, you probably shouldn't be asking me about what I think about something I read about Eric Clapton.

[jimbogumbo]

The paper got picked up by anti-vaxxers and that caused the brouhaha. A critic noted that  the author himself was an anti-vaxxer.

Competent editors would have had the afterthought excised.

As I've often claimed, the sociology of science is much more interesting than the philosophy of science!

Here is a fuller explanation of the funding for the study, which definitely came from an anti-vaxxer: https://www.truthorfiction.com/covid-vaccine-study-retracted/