The entire field of Metascience (see https://en.wikipedia.org/wiki/Metascience for a basic introduction) exists to examine the question of whether people who think they are doing science are really doing Cargo Cult science.
Their conclusion? Most ARE doing Cargo Cult Science. Particularly those working in psychology and medical research.
Let that sink in. Because it really, really matters. And given that it is true, perhaps you should stop objecting to the characterization and start thinking about what would be required for that to no longer be true.
Now you said, If you think that more research dollars should go to the second group (and I would likely agree), cool. Well, that is a mild version of what I am saying. I am saying that the decades long refusal of the field to fund or allow publication of any alternative lines of research is a genocidal crime. I mean that quite literally. They wasted billions of dollars entrusted to them and caused millions of unnecessary deaths.
This is not to say that amyloid research should not have been pursued. It was the leading hypothesis for good reason. But doctors have learned the hard way the importance of giving both a diagnosis and a differential. And then to look for evidence for the differential.
Researchers need to be better at doing the same thing. And as long as they are apathetic or, as you have been, openly hostile to it, they will continue doing Cargo Cult science. No matter how impressive the procedure looks, they are failing to do the basic things that are likely to make it work.
Straw man. I made no assertion about the the prevalence of cargo cult science in Medicine or any other field. I, too, have read Ioannidis. And Smolin. And Goldacre. And Kuhn. And Popper(well, some Popper. His writing is pretty dense.) Hell, I've even read Robert Whitaker. My point continues to be that mere use of correlation as evidence of cum hoc ergo pro is neither necessary nor sufficient for a charge of bad (or cargo cult) science because cum hoc ergo pro is a logical fallacy.
Your reference to Metascience is a red herring. Ionnides didn't conclude that basic logic errors were the problem: his focus was on reproducibility and statistical errors as well as problems in experimental design (probably publication of negative results as well, I can't remember). Further, 'Why Most Published Research Findings Are False' was published fifteen years ago. It's mainstream science now, and researchers are entirely aware of it. Feel free to charge (with evidence) that it's being ignored in experimental design, but it's certainly neither controversial nor obscure. It shows up in High School science curricula. Hell, I've used it in freshman college maths.
Feynman and Ionnides focus on experimental design, evidence selection, researcher bias or incompetence in selecting results, and publication of results. They don't say that you can't use correlation as evidence: the problem is not that there's too much evidence being accepted, but that there's too little.
Had you opened with
- "they're ignoring evidence they're wrong" or
- "this paper has twice failed in replication attempts" or
- "the study has too few subjects / no control group / too much noise to accurately measure these weak results" or even went all Smolin with
- "the entire research program is based on inadequate and uncritical investigations into a poorly defined and contradictorily applied hypothesis",
I'd be merrily mashing the up arrow. But you opened with "The amyloid beta hypothesis is the result of a failure to understand that correlation does not imply causation." You're asserting that hundreds of thousands of qualified medical researchers don't understand a fundamental principle of the scientific method.
Somewhat amusingly, the published results that prompted this particular discussion is good evidence against the plaque hypothesis (well, it will be after replication and a supporting metastudy). From a well-regarded researcher. I guess they're not all participating in the genocide.
I'm not "hostile" to the idea that AD research might have been misguided or even criminally wasteful of resources. I'm hostile to the assertion that the very close correlation between plaques and AD should not be guiding research and that it is clearly not only wrong but can be easily so demonstrated.
Here is the thing. The message from medicine has always been, "Well yes, there were problems in the past, but we do better now." And then they don't. A new round of problems shows up, gets a new round of criticism, a new round of, "things are better" happens, and the same old thing crops up. This pattern has literally held for centuries. Why should we believe them now?
Here are some examples. When I was growing up in the 1980s, we were taught ancient history like how much doctors resisted the idea of handwashing. Those who were curious could learn about the struggle to get double blind studies, and past disasters like the adoption of frontal lobotomies, or the development of hyper-radical mastectomies based on the theory that cancer spread slowly through tissues. (Breast cancer cells actually get metastasized through the blood.) I was so curious, and learned that we do better now.
Except that in the 1990s we found out that millions had had false memories implanted through bad therapy, and the idea of evidence based medicine became mainstream, clarifying how much better we didn't actually do in the 1980s. Under Bill Clinton we even briefly had a committee created to identify the most egregiously ineffective treatments to save costs. That committee's first recommendation was that except in obvious critical trauma cases, back surgery be a last resort to be tried after every other treatment option. The reason why was that it was the most expensive option, was the least likely to be effective, and long-term outcomes were worst.
Sadly, back surgeons were more effective at politics than at treating back pain. (They are also very profitable for hospitals.) Therefore that was also the committee's last report, and it sank without a trace. According to https://www.beckersspine.com/spine/item/45273-surgeons-perfo... reports we do roughlyt 1.6 million back surgeries per year, despite the fact that we have long had evidence indicating that we should do fewer of them. Let's state that a different way. Each year we spend billions of dollars on making the lives of over a million people worse because there is no willingness to act on data that has been available for decades.
THAT is the state of modern medicine.
Oh, and the much hyped move towards evidence based medicine? Every doctor will tell you that we now engage in evidence based medicine, the bad old days are gone. But estimates are that only half of medical procedures that we do have evidence of effectiveness. See https://www.vox.com/the-big-idea/2017/12/28/16823266/medical... for a popular article on that. It also shows that back surgeons have a lot of company. I guess things are better. We have some idea of how much current medical practice is useless..we just aren't doing anything about it.
Now do you see why I don't accept that things have fundamentally gotten better?
Now back to Alzheimer's.
Until less than 3 years ago, the field was openly hostile to research and researchers that didn't accept the amyloid hypothesis. This changed not because the field internally realized that it was doing anything wrong, but because of criticism received from outside the field. In Nature no less. So despite their theoretical awareness of how to do science, they failed to connect that to their own practice.
How many other fields are making similar mistakes but simply haven't been called out on it? My best estimate is, "A lot."
And yes, I did overstate my case in the passage you quoted. A more accurate statement is, "The unquestioned dominance of the amyloid beta hypothesis is the result of a failure to understand that correlation does not imply causation." As I have said repeatedly, there was nothing bad about following up on lines of evidence suggesting that amyloid beta was a good lead. The problem was rejecting all research into ideas that might question the favorite theory.
You know some other examples of where medicine went astray because people involved in a field rejected research into ideas that might question practitioners favored theory? Let's see. Frontal lobotomies, radical mastectomies, repressed memory therapy, back surgery, knee surgery, and so on.
At what point in this list should we start to suspect that there might be a pattern?
Their conclusion? Most ARE doing Cargo Cult Science. Particularly those working in psychology and medical research.
Let that sink in. Because it really, really matters. And given that it is true, perhaps you should stop objecting to the characterization and start thinking about what would be required for that to no longer be true.
Now you said, If you think that more research dollars should go to the second group (and I would likely agree), cool. Well, that is a mild version of what I am saying. I am saying that the decades long refusal of the field to fund or allow publication of any alternative lines of research is a genocidal crime. I mean that quite literally. They wasted billions of dollars entrusted to them and caused millions of unnecessary deaths.
This is not to say that amyloid research should not have been pursued. It was the leading hypothesis for good reason. But doctors have learned the hard way the importance of giving both a diagnosis and a differential. And then to look for evidence for the differential.
Researchers need to be better at doing the same thing. And as long as they are apathetic or, as you have been, openly hostile to it, they will continue doing Cargo Cult science. No matter how impressive the procedure looks, they are failing to do the basic things that are likely to make it work.