3 ms·
I feel like this study is missing the forest for the trees. Which section of society has the highest trust in scientists? Probably it's other scientists, closel
by mike_hearn 5y ago
I feel like this study is missing the forest for the trees. Which section of society has the highest trust in scientists? Probably it's other scientists, closely followed by science journalists. Yet this article is set in the paradigm in which people are divided into "the public" and "scientists", with the latter being universally correct and the former in need of education by the latter.
Over the past few years I've read a lot of scientific papers outside my normal field of reading (computer science/cryptography). My guess is over 100 in the past two years alone, or more if you count scan reading or checking only particular sections to follow up on a citation. The overwhelming impression this left is that many research fields are actually overrun with pseudo-science and that the institutions cannot detect this (or can but prefer not to).
This new study is especially ironic because social psychology is one of the biggest offenders when it comes to spreading misinformation. The entire field is rightly seen as junk due to the prevalence of non-replicable studies on 35 undergrads being generalized to the entire world, which are often then used to demand major political change. This particular study probably will replicate, because the claim it's making is so obvious as to be near tautological, but that is hardly better.
A good example of academic pseudo-science is claims about bots that subtly manipulate people's politics via Twitter. There are around 10,000 published papers on this phenomenon, which actually does not exist at all [1]. Claims it does are mostly derived from an academic "bot detector" that is claimed to be 95% accurate, but which actually has FP rates of around 30%-40% i.e. it looks and sounds like science, but is actually pseudo-science. The people writing these papers know their claims are junk (for example they almost never give lists of claimed bot accounts), but don't care because it lets them manipulate governments. California and Germany have both passed laws based on these claims.
Moreover, the study claims to have relevance to "COVID-19 misinformation". The formal COVID literature is a disaster zone. It's absolutely overrun with pseudo-science to an unbelievable degree. If you'd tried to tell me about the scale of the problem two years ago I'd never have believed you, so I won't be surprised if you don't believe me now. The sheer extent of pseudo-science in public health is a bit like the Matrix. Nobody can be told about it: really, you have to see it for yourself. Nonetheless, here are a few examples.
In [2] the very first sentence of the paper is a claim about public statistics that's factually false. Nobody seems to have noticed this despite that the entire paper is premised on the false claim, comes from Imperial College London, has 19 authors and was peer reviewed and then published in Nature. Having started with misinformation the paper goes downhill from there and makes several other pseudo-scientific claims.
In [3] the same team presents a model that is used to "prove" that lockdowns saved 3.1 million lives. The methodology is circular: the model by construction could not conclude anything else because it took as axiomatic that only government interventions can end epidemics (i.e. it assumed all epidemics have a single wave). The authors know that isn't true: the paper itself explains that it is "illustrative only" and that "in reality even in the absence of government interventions we would expect Rt to decrease
and therefore [we] would overestimate deaths in the no-intervention model", i.e. the authors admit that it's a fictional illustration of a world that isn't real and that its conclusions are wrong by an uncharacterized amount. Yet scientists don't care: this pseudo-scientific paper has been cited over 1300 times.
In [4] the authors claim that that model is "validated", which they define as: the model outputs roughly match the results from other models, not reality. This is a circular, pseudo-scientific definition of validity.
It's not hugely surprising that this stuff happens. There are strong social conventions in science that you always assume good faith on the part of other scientists, often to an absurd degree. In [5] an obviously Photoshopped/airbrushed image got peer reviewed and published in a medical journal. When someone pointed it out, the authors claimed that the patient merely wore the same shirt six months later, and the journal accepted this explanation! Anyone with their brain switched on can see it's the same image, but until the journal was mocked and embarrassed in public "science" was defending an obvious fraud.
So before social psychologists start lecturing the public on how they should be better at detecting pseudo-science, how about they clean up their own house first? Once academics are reliably detecting and flagging pseudo-scientific papers in their own journals, then and only then should they expound on the dangers of misinformation.
[1] https://blog.plan99.net/fake-science-part-ii-bots-that-are-not-c66129e5e3f5 https://blog.plan99.net/fake-science-part-ii-bots-that-are-n...
[2] https://dailysceptic.org/2021/08/24/examining-the-latest-paper-from-imperial-college-london/ https://dailysceptic.org/2021/08/24/examining-the-latest-pap...
[3] https://nicholaslewis.org/did-lockdowns-really-save-3-million-covid-19-deaths-as-flaxman-et-al-claim/ https://nicholaslewis.org/did-lockdowns-really-save-3-millio...
[4] https://github.com/ptti/ptti/blob/master/README.md https://github.com/ptti/ptti/blob/master/README.md
[5] https://scienceintegritydigest.com/2019/11/11/photoshop-as-a-management-modality/ https://scienceintegritydigest.com/2019/11/11/photoshop-as-a...