Hacker News new | ask | show | jobs
by derefr 1819 days ago
> But, yeah, as I said, psychology is pseudo-science.

Science is about doing experiments to get data that allow you to create+refine models of reality that make predictions on what further data will look like.

Psychology is a science. People may argue whether it is a hard science, but it’s doing all the science things.

What is the difference between an RCT on how a drug affects cancer (given some formal rating scale for cancer), vs. an RCT on how a drug affects ability to concentrate (given some formal rating scale for ability-to-concentrate)? The former is considered medical research. The latter is considered psychological research.

> Well, I don't expect a psychologist to have enough expertise for a biopsy, but could I expect to at least have a couple holes drilled in my skull, or maybe some electroshock therapy?

Keep in mind that psychologists — i.e. people who do academic research in psychology — are drilling holes in skulls all the time. Y’know, on rats. (And not just on dead rats. They’re often implanting wires and such.)

Still, though: why? A brain biopsy is almost-always worse/higher-risk than just putting up with whatever was wrong with you before.

Also, I don’t want people to drill holes in my skull. Most people don’t. It is, in fact, considered unethical by most medical boards to drill holes in a patient’s skull, if what you’re treating for would not be worse than a hole in the skull. (And a hole in the skull is very risky, in terms of liability to infection, stroke, etc.)

This is my point: psychiatrists are people who, like IT help desk techs, try to diagnose a thing by hearing it described over the phone, with no ability to touch or interact with it. Psychology is the model, the best set of predictions we’re been able to attain, for how the mind works, given that we can only interact with it this way.

Psychologists try very hard, using a lot of rigor and very powerful statistical methods, in an attempt to extract signal from the super-noisy clinical input of the practice of clinical psychiatry and of human psychiatric research. (Plus animal psychological studies, where we have the alternate problem of trying to model a mind we can probe directly but can’t communicate with.)

“Unethical psychology” would be a hard science indeed.

> When you say 'our best understanding' you mean a psychologists best understanding?

I mean humanity’s best understanding. The academic-scientific ‘us’ — everyone working together to advance the frontier of knowledge.

—————

Addressing your comments as a whole, you seem to have conflated the practice of clinical psychiatry, with the medical science of psychology.

“Psychology” is just what neuroscientists call their neurological behaviour studies, when the study doesn’t involve or rely on a white-box model for what’s happening, only a black-box behavioural model.

In modern practice, there are no psychologists who aren’t neurologists; no psychology paper is being written by someone who isn’t a neuroscientist. “Psychology” is to “neuroscience” as “ML” is to “Computer Science” — i.e. a specific sub-discipline that some researchers might focus on, but not because they lack the skills outside of that discipline; rather only because they enjoy the process of doing that particular type of research more.

Psychiatry is the practice of using psychological findings in a clinical, medical context. Psychiatrists are doctors, who have then further specialized by learning deeply+broadly about the various models-of-understanding that psychologists have developed. They know as much about medicine as any other doctor; they just have the additional understanding that e.g. “depressed people aren’t just sad.” (Which is, y’know, something we had to prove, and all the papers that do that are psychology papers.) Or “being gay is not a disease.” (Which, again, something psychologists had to prove.) Sadly, you can’t rely on a non-psychiatrist doctor to be aware of these sorts of things.

As such, psychiatrists are probably the doctors it’d be most beneficial to talk to, if you have a problem that is potentially psychologically rooted. A regular GP, who never touched any of that specialty while getting their degree, will be able to recognize organic diseases, but won’t necessarily recognize psychiatric syndromes, and so will be very likely to mis-diagnose a purely-psychiatric syndrome as an organic disease.

1 comments

> Psychology is a science. People may argue whether it is a hard science, but it’s doing all the science things.

True, when I say 'science' I mean 'hard science'. Every theory is basically an arbitrary statistical model trained on extremely noisy signals. Generally, it can't carry out controlled experiments or predict new phenomena.

> What is the difference between an RCT on how a drug affects cancer (given some formal rating scale for cancer), vs. an RCT on how a drug affects ability to concentrate (given some formal rating scale for ability-to-concentrate)? The former is considered medical research. The latter is considered psychological research.

You can measure the size of the cancer in a fairly direct and objective way. To measure ability to concentrate you need to have a statistical model that interprets a very noisy signal.

> Also, I don’t want people to drill holes in my skull. Most people don’t. It is, in fact, considered unethical by most medical boards to drill holes in a patient’s skull, if what you’re treating for would not be worse than a hole in the skull.

It was a joke about lobotomies...

> “Psychology” is to “neuroscience” as “ML” is to “Computer Science”

It's more like what 'astrology' is to 'astronomy'.