Hacker News Reader

Comment by aliasEli | Hacker News Reader

The main problem with the article is that it assumes that all deaths after vaccination are due to the vaccination, which is nonsense. If 100% of the people are vaccinated they would describe all deaths to it.

Also they do not seem to correct for age. We started the vaccinations in the Netherlands with the oldest persons, who were already much more likely to die. So the LAREB results are not representative for the entire population. They will give an overestimate of the mortality by vaccinations.

The Israeli study [0] uses a large sample of people over 16. However they explicitly excluded persons bound to their home or were nursing home residents. This will give a low estimate for the effect of vaccination because the sample includes many healthy younger people, and excludes the highly vulnerable ones.

[0] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7944975/

View on HN

I think the important story here isn't that some controversy addicts/anti-vaxxers wrote a bad paper, but that the review process at a journal called Vaccines let trough a 4chan level analysis. And this during a world wide drive to vaccinate everybody no less!

Sadly it's too hard to attribute the deaths via vaccine hesitancy that this paper will cause, so none of the reviewers can get charged with involuntary manslaughter.

If you dig deep enough, many of the peer-reviewed papers are unreproducible.

In some fields sure, and whenever you have small sample sizes you'll unavoidably have many contradictory results (unless everyone agrees on priors beforehand), but I don't think vaccines is a field with a replication crisis comparable to psychology.

Medicine is absolutely a field full of bullshit papers that misuses methods and statistics. You'd think they would be better than psychology but not really, they are often even worse at math and statistics than the psychology papers. You'd have to go to biology papers to get reliable rigor, they do a lot on vaccines.

Manslaughter for publishing a bad scientific paper? Have you lost your mind?

We don't want to contribute to vaccine hesitancy, but we also don't want to adopt a totalitarian worldview where any deviation from "yay vaccines!" is tantamount to murder.

For the reviewers? No. But if we find out people in the journal hand picked reviewers likely to pass the papers in order to see it published, well, I'd like to see at least a civil suit against them. That's not science when you load the dice, it's fraud.

This kind of retribution mindset doesn't help. Will scientific progress increase when we get more lawyers involved? Is that what society needs? No.

Science should be a process where eventually truth wins out. For this to happen scientists need to be able to publish their findings without fear of retribution. They already risk losing funding, they already have poor job security, they already need to meet a quota for papers and citations. Adding more negative incentives on top of this won't do society any good.

Bad reviewers are extremely common. I wouldn't read too much into lazy people not double-checking the stats carefully.

The one highly-enthusiastic reviewer is a bit suspicious, but I'd more suspect it was a case of the author recommending a friend as a reviewer (a common practice). But it could also just be by luck they got a person who was negative towards covid vaccinations so wanted to be very supportive.

Isn't peer review usually blind?

It's very common for journals to ask for reviewer suggestions. For example I guess this is the journal in question:

https://www.mdpi.com/journal/vaccines/instructions

"Reviewer Suggestions

During the submission process, please suggest three potential reviewers with the appropriate expertise to review the manuscript. The editors will not necessarily approach these referees. Please provide detailed contact information (address, homepage, phone, e-mail address). The proposed referees should neither be current collaborators of the co-authors nor have published with any of the co-authors of the manuscript within the last five years. Proposed reviewers should be from different institutions to the authors. You may identify appropriate Editorial Board members of the journal as potential reviewers. You may suggest reviewers from among the authors that you frequently cite in your paper."

Obviously people are supposed to do so ethically, but cabals of friends can potentially hack the process by suggesting each other and working together.

God, the current publication system is worse than useless....

It depends on the journal. Anyhow it is probably useless anyhow because most scientist refer very often to their own work or that of their research group. If you have some knowledge of the research field, it is usually not too difficult to guess who the authors are.

Referees are normally anonymous. But once they start advising to insert some references to some obscure and/or irrelevant articles...

sort of, but many reviewer communities are small enough that you infer the reviewer's identify from the papers they ask you to cite and the argumentation style.

Manslaughter is not tantamount to murder, that's the entire point of the concept!

Further, your characterisation of the article as a mere "deviation" is grossly inaccurate: their text is not a honest look at vaccine dangers but rather it maliciously picks data to get the worst result misuse can produce. The only thing they shy away from is outright fabrication.

> Manslaughter for publishing a bad scientific paper? Have you lost your mind?

I recommend you get acquainted with the definition of manslaughter, otherwise you wouldn't complain or object to it after learning the basics of this particular paper and how it's broken logic is being abused as a denialist propaganda tool.

For future reference, here's Cambridge Dictionary's definition of manslaughter:

> the crime of killing a person when the killer did not intend to do it or cannot be responsible for his or her actions:

https://dictionary.cambridge.org/dictionary/english/manslaug...

> a journal called Vaccines

I'm an academic mathematician. Your pointing out the name of the journal is an interesting point.

Pop quiz: out of the journals Aequationes mathematicae, Indagationes mathematicae, Inventiones mathematicae, Quaestiones mathematicae, three are mediocre and one is absolutely top-notch. Without googling can you guess which?

MDPI is a publisher that was started recently, has apparently tried to grow very fast, and engages in questionable practices. I peer-reviewed for one of their journals once and had a very negative experience. Never. Again.

They've also started journals with good sounding names. For example there was no mathematics journal with the title "Mathematics" so they started one.

I've instructed my mail client that *@mdpi.com is to go straight to spam. I got invited to be on the editorial board of one of their journals, and I felt no guilt about not responding. Not too surprised to see them publish crackpottery.

I think this is a fascinating question. There does seem to be an association between 'better' names and higher quality, and I've been wondering why that might be. Older journals have a wider range of available names so choose better? Association only in our heads between the journal quality and its name subsequent to reading the content? (I got your pop quiz completely wrong; would a junior mathematician who hadn't read or heard of any of them think that 'inventiones' was inherently a better name?) Or is a 'better' name likely, ceteribus paribus, to attract better manuscripts - starting a feedback loop that drives it up? I can't help thinking that a journal called 'Mathematics' has to have better prospects than the average questionable new journal.

My own field is political communication. One of our leading journals is 'Political Communication', which could exemplify any of the above. But another is 'The International Journal of Press/Politics', which should by its name be complete shit and turns out to be excellent.

Hm, Science describes it as a "reputable open-access journal" which based on your account I now must assume to be low key gloating.

Certainly this debacle should lower their rating to "low repute journal".

I think mostly journals like this claim to be "reputable" based on their impact factor rather than based on any particular claims to accuracy or quality of published research.

Out of curiosity, are there notable journals for mathematics that are published by MDPI?

Not to my knowledge, no.

> Pop quiz: out of the journals Aequationes mathematicae, Indagationes mathematicae, Inventiones mathematicae, Quaestiones mathematicae, three are mediocre and one is absolutely top-notch. Without googling can you guess which?

Is the good one Indagationes ?

(Edit: looks likely. I picked it because it seemed the better name)

Bzzzt. :)

Yeah, that's the only one whose meaning you can't guess if you don't read Latin. Inventiones, as it turns out.

Native Portuguese speaker here - "indagações" (pronounced pretty close to the Latin), is an infrequently used word meaning asking questions ;-)

Seems the Italians kept it around in the original spelling, too.

P.S. Hey, it's English, just barely: https://www.merriam-webster.com/dictionary/indagate

Gosh, I'm tired of MDPI emails asking for reviews in 10 days, or asking for papers, etc. They are terrible.

I also made the mistake of reviewing for them once. It was a very easy review, as the paper (also mathematics) was absolute crap, and I said as much. It was published anyway, and now I'm definitely on a list. My spam box slowly accumulates more and more MDPI emails.

'Correlation is not Causation' is one of the most basic modern scientific principles, yet is ignored here by the authors and the reviewers.

By the same reasoning, how many deaths and illnesses would we attribute to people getting photographed with a mobile device, or drinking refrigerated water - i.e., how many times each day is someone photographed, or drinks cold water, who also falls ill or dies within a week? We'd have to ban cameras and refrigerators...

This entire episode is mostly evidence supporting Einsteins statement that "Two things are infinite, as far as we know - the universe and human stupidity; and I'm not sure about the universe."

I disagree. The main problem is the way they calculate NNTV as if the risk of infection was linear. It's not since there is a positive loop. In the absence of a vaccine, rate of infection grows which in turn increases the risk of infection.

> The main problem with the article is that it assumes that all deaths after vaccination are due to the vaccination, which is nonsense

the main problem with COVID deaths is they are counted as COVID deaths when COVID is basically what pushed already very sick people over the edge. Ever wondered why the large majority of COVID deaths were among people with numerous co-morbidities and already in advanced age?

If you want to make that argument you have to make both ways, and not cherry pick the data.

But that's just easy to verify. We can take the 5-year average and compare that to 2020/2021 deaths. In countries such as Belgium the "excess mortality" corresponds to the reported COVID19 deaths.

you still get this excess mortality if covid merely brings the deaths forward by 3-6 months.

please show me your sources.

Reply on news.ycombinator.com