Interviewer: It is striking how scientific studies, also in this corona crisis, reveal very diverse results. Based on these results, scientists can defend almost diametrically opposed theories as the only truth. How is this possible?
Mattias Desmet: The research on corona is indeed brimming with contradictions. For example, regarding the effectiveness of face masks or hydroxychloroquine, the success of the Swedish approach, or the effectiveness of the PCR test. Even more curiously, the studies contain a huge number of improbable errors that a normally sane person would not be expected to make. This is still the case in terms of establishing the absolute number of infections, while a schoolchild knows that this means nothing as long as the number of infections detected is not compared with the number of tests taken. Obviously, the more tests you carry out, the more likely your infection rate will increase. Is this so difficult? In addition, it should be kept in mind that the PCR test can yield a large number of false positives, because the technique is widely misused for diagnosis. Together, this means that the inaccuracy of the figures distributed daily by the media is so great that some people understandably suspect a conspiracy, albeit apocryphally, in my opinion.
Again, this phenomenon is better placed in an historical perspective, because the problematic quality of scientific research is not a new issue. In 2005 the so-called “replication crisis” erupted in the sciences. Several committees set up to investigate scientific fraud cases found that scientific research is teeming with errors. Often the stated conclusions are of very dubious value. In the wake of the crisis, several papers appeared with titles that leave little to the imagination. In 2005, John Ioannidis, Professor of Medical Statistics at Stanford, published Why most published research findings are false. In 2016, a different research group wrote about the same topic, in Reproducibility: a Tragedy of Errors published in the medical journal Nature. These are just two examples of the very extensive literature describing this problem. I myself am well aware of the shaky scientific foundation of many research results. In addition to my master’s degree in clinical psychology, I earned a master’s in statistics. My doctorate dealt with measurement problems in the field of psychology.
– Patrick Dewals: The Emerging Totalitarian Dystopia: An Interview With Professor Mattias Desmet
Original text here.
My dad was a very good statistician.
He frequently told me that 99% of everything is shit.
Was he wrong?
How’s this for a proposition:- Outside of the Social Sciences, there is a lot of research projects that only have one goal, to harvest government research grants. Since the purpose is to get the grant money, once they’ve got it, the ‘research’ is a secondary matter.
The research in any field that has a significant biology component is suspect. The experimental subjects (rats, people…) are so highly variable that sample sizes need to be enormous.
The famous Nurses Health Study is a longitudinal, life-long monitoring of the health of some 275,000 female nurses. And what does it tell us: something about white, college-educated, middle class women.
Some of its results might be relevant to the lives of men, or working class people, but as big and as long-lasting as the study is, its results are very limited.
I used to tell my students that environmental research was so hard it was easy. No one could ever check your results.
“Outside of the Social Sciences, there is a lot of research projects that only have one goal, to harvest government research grants.”
Why outside of the Social Sciences?
Because inside the social “sciences”, it’s 100%?
I have personal experience, back in the Sixties. When my advisor gave me my master’s project, he showed me the thesis my predecessor had written. One graph showed he’d taken a lot of data points. It turned out he’d taken the data points at the beginning and the end, drawn a straight line between them, and prettied it up with dots in between. “Did you really take all those points?” his advisor asked. “No, I just thought it looked better with dots.”
They gave him his Master’s degree and tossed him out of the doctoral program and the university. I did it much more carefully, and got a degree and an article accepted by Nuclear Instruments and Methods.
These things happen. Sometimes it’s laziness or incompetence.
Every year, there are numerous studies reporting that a proposed drug is ineffective or has unacceptable side effects. There are many more such studies than studies reporting effectiveness and safety. (That’s why the “mortality” of projects is about 80% at Phase II and again at Phase III.)
Are most of these studies false?
Can’t remember where I read it – it may have even been here on Samizdata – but “putting the word ‘social’ in front of the word ‘science’ is a modifier not unlike putting the word ‘witch’ in front of ‘doctor’!”
Money, always money..
money and medals..
money,medals, and awards..
money, medals, awards and tenure
Similar to putting the word climate before the word scientist.
Many years ago I remember sitting in a public bog where, instead of a roll of bog paper there was a metal box with a slot in it from which you could pull individual squares of paper to clean your bott. Above said box, some wag had written, sociology degrees please take one.
Fred Z. one of the books I read in my youth was called “95% Is Crap” by Terry Arthur – it is a fine book and it explains how all the political parties and interest groups of the early 1970s were, well, talking Crap – they had about as much knowledge of basic economics (or anything else) as the vile creatures who write the Economist magazine today – whose criticism of the British budget was that it did not increase the rates of income tax and increase welfare payment levels.
I am not making that up – to the “free market” Economist magazine income tax rates should be increased (not just allowances not increased – which is a tax increase they did not even mention), and welfare levels are too LOW.
In a world where the leading “free market” magazine is a Corporatist Big Government rag like the Economist is it any wonder that the establishment lies about HEALTH not “just” economics.
“But health is human lives Paul” – so is economics. Economic collapse will kill vast numbers of people around the world.
Let us say there was an honest difference of opinion about the effectiveness of Early Treatment for Covid 19 – let us say that some people honestly believed that (for example) hydroxycholoroquine and ZINC and azihromycin were ineffective – but why would they SMEAR the treatment? Why pretend that it was about “poisoning people” trying to “kill them”?
And what about vitamin D3 – what possible HARM could it do to increase vitamin D3 levels in the general population? Let us say (for the sake of argument) that this would be ineffective – but why work to PREVENT that happening?
No, an honest disagreement, does not fit the facts – a lot of powerful interests did not WANT the death rate reduced.
Think about that – think about just how terrible that it.
The “mainstream” media and the bureaucracy (both governmental and corporate) is under some very bad influences.
When I was in Oxford, I occasionally helped a friend in the biological sciences who, though no statistician, sometimes spotted very elementary, very obvious (so obvious my unmathematical friend spotted them and just came to me to check) incompetence (incompetence, not – that I could see – motivated bias) in applying statistical tests – in papers that had been successfully published by high-up academics. There were biology research groups that understood their need of high-powered statistical review – and there were others that did not know how much they did not know. It was worse in the social sciences.
Honest (albeit slovenly) statistical mistakes are far from the whole story but they are, and to some degree were, part of it.
79 percent of the population know that 67.2 percent of statistics is made up. Or is it the other way around?
Paul Marks –
Yes, like that disastrous Lancet article on HQC that had to be pulled a month after it was published due to obvious egregious errors in the patients selected and the almost fatal levels of HQC administered without either Zinc or Azithromycin. That article was treated as THE defining study proving that the HQC trio was worse than useless across the MSM and social media, while they suppressed other studies showing clear lifesaving benefits when used properly. The completely unnecessary death numbers were used to justify the economy destroying lockdowns. The booming economy of 2019 would have guaranteed Trump’s re-election. Therefore the economy HAD to be crushed along with justifying massive mail-in ballots across the country.
It. Was. Deliberate.
George Atkisson – yes. President Trump had mocked them, the World Economic Forum and so on, to their faces – and so he had to go. And if half a million Americans “had to” die to achieve that objective – well they did not really care about that. Although I think it was about more than President Trump.
As the saying goes “it is nothing personal” (although I think it was partly personal – Donald John Trump had, I repeat, mocked-them-to-their-faces and the World Economic Forum Corporate establishment do not tolerate that), “it is just business” – in this case the business of creating world wide totalitarianism “Stakeholder Capitalism” (the Corporate State) in the name of “Sustainable Development”.
The World Health Organisation is still paying for ads (on Google) and so on, trying to convince people NOT to engage in Early Treatment.
That is correct – your tax Dollars Mr Atkisson are being sent (by Dr Fauci) to the World Health Organisation – so they can be used on ads to enrich Google (one of the most despicable of the international mega Corporations) and to INCREASE the death rate of Covid 19.
When did the Federal Government lose any sense of honour, any basic honesty?
Cynics say it was always as it is now – but I do not believe that.
I believe that, for example, the Justice Department and the FBI still had some sense of decency as late 1992 – before Mr Clinton came in and corrupted things.
Of course I may be seeing the past via rose tinted spectacles – perhaps, for example, the FBI was always as sickeningly corrupt as it is now, persecuting the innocent and covering up the crimes of the guilty.
As for today…..
I have some trust in ASIO – the Australian Security and Intelligence Organisation. I do not think they are the despicable people that the American agencies have proved themselves to be.
Perhaps an Australian will now come in and point out how touchingly innocent and dumb I am.
Australia has gun control and restrictions on Freedom of Speech – but at least being “fitted up” by the state (prosecuted for crimes you did not commit) is not the norm there, and at least one does not have to watch people guilty of terrible crimes (such as the vile creature Joseph Biden – at least he was a vile creature before he lost his mental capability) installed as the head of the government – rather than go to prison as they deserve.
The motto of the modern FBI really does seem to be – persecute the innocent and cover up the crimes of the guilty.
And as was revealed yesterday – Dr Fauci and co take their orders from the People’s Republic of China, it really is as bad as that.
@FredZ
I’m not sure about 99%. Let’s say 85% crap, 10% fraudulent, 3% not very interesting, 1% quite important and 1% important.
Science (good science) is hard and usually time consuming. It doesn’t necessarily need large grants or masses of equipment (although it may).
My father, also a statistician, worked closely with a very eminent biomedical researcher. They used to regularly go through the latest issues of the key journals with the biomedic instantly dismissing about 98% of the articles as rubbish.
check out the song, “Statistician’s Blues”, by Todd Snider. It explains everything.
I would suspect that the track record of various kinds of applied science would have a better track record. If you need to get stuff right in order for it to actually work you can’t really afford to be sloppy.
Clinical trials don’t show any effect from HCQ.
Some early studies said that it did, but they were small observational studies. Randomised large scale trials find no benefit to HCQ.
https://www.nejm.org/doi/full/10.1056/NEJMoa2022926?query=featured_home
HCQ group, 1561 patients. 27% dead in 28 days. 60% discharged in 28 days. 31% needed ventilation.
Control group, 3155 patients. 25% dead in 28 days. 63% discharged in 28 days. 27% needed ventilation.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7471562/
Meta-analysis of included studies revealed no significant reduction in mortality with HCQ use[RR 0.98 95% CI 0.66–1.46], time to fever resolution (mean difference − 0.54 days (− 1.19–011)) or clinical deterioration/development of ARDS with HCQ [RR 0.90 95% CI 0.47–1.71]. There was a higher risk of ECG abnormalities/arrhythmia with HCQ/CQ [RR 1.46 95% CI 1.04 to 2.06].
https://pubmed.ncbi.nlm.nih.gov/33617702/
This rapid review appraises available evidence on the viability of vintage antimalarial drugs chloroquine (CHQ) and its analog hydroxychloroquine (HCQ) repositioned for SARS-CoV-2 prophylaxis and COVID-19 treatment. Findings suggest neither the use of CHQ nor HCQ singularly, or concomitantly, with azithromycin and/or zinc provide definitive benefits for use against SARS-CoV-2 infection or COVID-19 illness. Moreover, administration of these medications was linked to significant and sometimes fatal complications.
https://pubmed.ncbi.nlm.nih.gov/33247380/
Clinical recovery after 28 days was achieved by 79.2% in the zinc group and 77.9% in zinc-free treatment group, without any significant difference (p = 0.969). The need for mechanical ventilation and the overall mortality rates did not show any significant difference between the 2 groups either (p = 0.537 and 0.986, respectively). The age of the patient and the need for mechanical ventilation were the only risk factors associated with the patients’ mortality by the univariate regression analysis (p = 0.001 and < 0.001, respectively).Zinc supplements did not enhance the clinical efficacy of HCQ.