Hacker News Reader

Comment by hlieberman | Hacker News Reader

A retracted, un-peer-reviewed article published by an economist and a medical doctor, neither of whom have any experience in public health.

Might as well be written by Andrew Wakefield.

View on HN

I can quickly identify some potential issues with statistical controls and low numbers of samples, but a dismissal based on credentials is a weak logical fallacy called credentialism. I see it everywhere, especially on forums of people who are likely to have and value credentials (HN). The authors' credentials (which are really not lacking anyway) have nothing to do with the merits of their argument. Especially on an internet forum, where there is no shortage of people with free time to analyze arguments.

I see your point, but the opposite is just as bad or worse. The idea that studding for years has no merit and that the opinion of a professional with 10 or 20 years in a field (20.000 hours - 40.000 hours) has the same value of the opinion of some dude on the internet that thinks he is really smart and googles shit for a couple hours, even if it's hundreds of hours, is just ridiculous.

The retracted part isn't a credential though - it shows that someone with authority/knowledge decided there was something really wrong with the piece.

The graphs are preposterous. This looks like an exercise in finding what you want to see in noise. I’m horrified at the thought of this paper being disseminated to the general populace that won’t understand that.

Hungary by far has the highest “excess mortality” but the lowest vaccination rate in the three (!) mass vaccination countries used for the whole paper and in Table A-1. In fact remove Hungary and there isn’t even an effect I assume. They are just averaging the three countries together! Israel and UK aren’t even high it’s just that Hungary is super high and doesn’t have lots of other countries to average out its variance. Czechia, Slovakia, and Poland have vastly higher excess mortality than Israel and UK but are in the low vaccination category. It’s nonsense.

Do you have any problems with the substance of the article?

I do. Look at the table A1 in the appendix. They separate the countries in two groups:

Mass vaccination countries: Israel, United Kingdom, Hungary

Moderate vaccination countries: All the other in Europe, or something like that.

To simplify, let's make a new column with the average of the excess of deaths, and sort them. The biggest ones are:

Hungary: 39.3%

Slovakia: 26.7%

Poland: 26.3%

Czechia: 25.0%

Bulgaria: 20.3%

Cyprus: 12.3%

Austria: 10.3%

United Kingdom: 10.0%

Israel: 8.3%

Other: 8.0% or less

Average: 4.8%

So clearly Hungary is a huge outlier.

Back to the three mass vaccinated countries:

If you compare Israel to the moderate vaccinated countries, the different is not statistically significant.

If you compare United Kingdom to the moderate vaccinated countries, the different is not statistically significant.

If you compare Hungary the moderate vaccinated countries, the different IS statistically significant. Moreover, it's much bigger.

The correct conclusion is that something extraordinary bad happened in Hungary, not that something extraordinary bad happened in the mass vaccinated countries.

You can at least look at the same data they did.

Reply on news.ycombinator.com