TMPHBITEU
What are the flaws in the scientific method?Originally Posted by mbdonner
PLEASE do not confuse Medicine with Science. They are absolutely not the same thing.I have always read articles and papers with a critical eye, does the methodology make sense, what merit or value do the conclusions have. Cui bono. Psychiatry has spent the last 50 years trumpeting the "chemical imbalance" theory, pushing SSRI's, but NONE of the research ever found results much better than placebo.
Having said all of that, the idea that anecdotes spread around the internet is better is exactly what you spend a lot of time debunking. Otherwise, you would be teaching quarter squats, because full squats are hard on the knees. That's the common "anecdote."
"Anecdotal evidence" is evidence based on the personal observations of the reporting individual. This makes the reporting individual quite important in assessing the veracity of the quality of the information. To cite a popular example, there are no "peer-reviewed studies" concerning whether the sun comes up in the east every day, and the sunrise/sunset data tables are not typically subjected to peer-review, because the men with clocks know how to use them, and we just trust them to do so. And, most importantly, it would be hard to obtain funding.
The first definition of "anecdote" is "a short interesting or amusing story." I am famous for those.
You forgot the peer-review part.I think you actually use the scientific method. You observe, you generate hypotheses, you test them in your lab, you adapt, adjust, publish, open yourself up to critique, others come along, test your hypotheses, publish.
Member
"There is no algorithm without an executable . . ." - Alexander Stepanov (https://www.linkedin.com/in/alexanderastepanov/)
(Might have posted this around here before - but it is powerful advice that is easily glossed over.)
TMPHBITEU
Not going to LinkedIn. If it is interesting, post it.
Member
I should have been more careful in my language, and said limitations, not flaws, in the scientific method. Setting aside fraud or errors, one of the limitations is the use of small samples to generalize to large populations. The other is that the choice of statistical analysis can result in misleading outcomes. Although not exactly a limitation of the method, the reporting of results can result in the public thinking something works when it doesn't. These issues can be compounded by the pharmaceutical industry. So for example, SSRI's. The data was always clear that very few people got much benefit from the medications. But there was big money to be made. The "chemical imbalance" theory was developed to promote the use of SSRI's. Although the limited benefits were readily accessible, pharmaceutical reps pushed the medications to psychiatrists who didn't really read the research. This spread to psychologist's in a kind of "word of mouth" way, and eventually it became the "standard of care" to refer patients with anxiety and depression for medications, even though there was little evidence to support their utility. When the patient didn't get better, they switched medications. Like with back pain, most people get better on their own.
So is that a limitation of the scientific method or the result of ignorance, greed, capitalism, all of the above?
TMPHBITEU
This is a silly question, and you answered it yourself.
Member
To find real "science", go out into the country and find an unemployed scientist living in a brick ranch who mows his own lawn. Listen to him. There's a reason he's unemployed.
That's how bad is it now. We are there.
Member
Stepanov is one of the best programmers in the world, now retired after a distinguished career. Implied in that career is long experience with impractical or impossible “paper designs” - one example being his extensive work and struggles on the C++ standards committee (ISO).
Some process, recipe, rule, or guidance might read perfectly on paper, but won’t compute when given to a computer - i.e. it violates some physics in a non-obvious way.
Member
That has nothing to do with Phillips' patent disclosures.
Member
I was referring to the issue of reproducibility in Satch's post, the trailing part of which was "...then being found not to work when implemented in the field...".
I haven't read Phillips' disclosures, so I do not know if they are "reducible to practice". If not, then there is a useful parallel to a putative algorithm with no corresponding executable.
It's only an abstraction - take it or leave it.
Starting Strength Coach
No. To find a real scientist, go the gym, clinic, workshop, or lab of a coach, clinician, inventor, craftsman, or research investigator of true integrity and independence, who brings his full intelligence, creativity, training, study, practice, and pragmatism to bear on the problems before him. There are such people. I know and have worked with many of them. To suppose they aren't out there anymore on the basis of fucking Stu Phillips and (admittedly) many other disappointments is to surrender to catastrophization, scientific nihilism, and a blind despair. It's not like we found out Niels Bohr fudged his equations. It's fucking Stu Phillips.
The corruption of human institutions and enterprise is not new. It is a perennial and messy problem, nothing more.
Forget science... Look at this industry! It crawls with stupidity, superstition, arbitrary ivory-tower navel gazing, silly bullshit, and outright malpractice.
AND YET: It is vibrant, growing, and strong. Look at the emergence of the profession of barbell coaching, a transformation due in no small part to the psychotic and abrasive savant who runs this board. The emergence of this new Discipline of Practice is messy, painful, all-too-slow, and better than the alternative. Corruption, decay, germination, growth. The world has never been without them all. We take a deep breath, demand better, and move on without losing our shit.
Posting Permissions
