Comment by timr | Hacker News Reader

Comment by timr | Hacker News Reader

Parent Full thread

HCQ doesn't work for Covid, but this is a fundamentally political retraction that won't accomplish anything.

If I had to line up Covid-related papers that should be retracted on the grounds that the science was shoddy this paper would certainly be on the list, but it would be on a list with other, absolutely horrible, irresponsible stuff related to (for example) masks [1], that the same people who railed against this paper continue to believe, almost as religion.

In the Covid debates "both sides" have their shibboleths that are based on atrocious science, so let's not get too sanctimonious about retraction of a single paper.

[1] Since people are downvoting because I said the word "masks", here is just one high-profile example: https://www.nejm.org/doi/full/10.1056/NEJMoa2211029

This paper was found, in subsequent analysis, to have cherry-picked their data sets to achieve a pre-determined outcome:

https://arxiv.org/abs/2307.11974

If you find yourself reacting emotionally to this particular topic (or any other related to Covid), you should take a step back and consider that you might be an example of what I'm talking about.

Masks aren’t a chemical inserted into your body. Their risk is practically zero. While some of the information around effectiveness may be a bit exaggerated, masks were a net benefit that Americans needed to learn about. Next pandemic feel free to forgo your mask as a core of the population has learned the advantages of masking and is fine with it.

> Their risk is practically zero.

In fact, no. We're finding out that children are having long-term speech and educational issues related to masks [1,2]. And that's without looking very hard, because people continue to make bald assertions about risk that deter actual, high-quality research into the question.

> masks were a net benefit that Americans needed to learn about

In science, claims must be proven. We don't just get to make things up.

[1] https://www.gov.uk/government/publications/evidence-summary-...

[2] https://www.theatlantic.com/family/archive/2022/03/how-masks...

These reports could be debunked in 4 years.

What horrible irresponsible stuff about masks are you referring to?

Literally hundreds of papers. Basically any paper published about masks during the pandemic was bad, save for a few RCTs. But I'll give you one particularly egregious example that comes to mind. The "Boston school mask study" in NEJM [1] claimed that mask mandates led to an almost 30% reduction in school cases during the pandemic. Subsequent reanalysis of the work -- including going back to the original source data -- showed that not only is that not true, the researchers blatantly cherry-picked their data, and also chose analysis time windows to optimize their desired outcome.

The original paper should be retracted. Full stop.

[1] https://www.nejm.org/doi/full/10.1056/NEJMoa2211029

[2] https://arxiv.org/abs/2307.11974

Maybe I missed it, but for [2], I'm not seeing any info about whether testing rates were consistent between schools with mask mandates and schools without. (Considering that tests were as politicized as masks, I don't know if there's a correlation between "didn't require masks" and "didn't require testing.")

> I'm not seeing any info about whether testing rates were consistent between schools with mask mandates and schools without.

The paper tells you that the Boston and Chelsea districts were outliers in terms of testing in the first sentence of the Results.

But it's essential to the claims of the original paper, so you should look there, and cross-apply the same criticism. If you can't prove that testing rates were consistent across schools, how can you make the claim of a 30% increase in positivity?

(Hint: Cowager et al. provide the data buried in a supplement. They show that testing practices, in fact, were neither consistent nor stable across time. Both violate the requirements for applying the statistical method used in the original paper.)

I think you're wasting your time arguing with a True Believer.

What are the "absolutely horrible, irresponsible stuff related to (for example) masks" that you speak of?

I just made a reply to a sibling comment with an egregious example. This is far from the only one.

If a paper advances false claims it should be retracted. What those claims are and who might find said claims politically convenient are utterly irrelevant. This is not a political issue. Attempting to frame it as such accomplishes nothing of value.

I agree, and I'm saying that what actually gets retracted is basically a function of the activism of people who are particularly motivated, which is a matter of politics.

I'd entertain evidence that supports your assertion. Do you have any?

Not an assertion; provided links to paper that needs retraction elsewhere in thread.

Beyond that, the article specifically says how this paper got retracted -- a group of activists pushed for it. I encourage you to push for that other paper to be retracted as well.

Or money.

Reply on news.ycombinator.com