The Replication Crisis in Psychology

[@silvertube52]

The "replication crisis" in psychology is a part of a broken academic system. Psychology has no funding. You can only have a career if produce many publications. You can only generate publications if you show positive results of a new and unusual findings. A principled researcher will more likely end up a cab driver than a professor. While working on my PhD in social psychology I observed researchers engaging in what we now call "p-hacking." One of my advisors repeated minor variations on the same study over and over, and published when it got statistical significance. I knew that was not valid and spoke openly of the problem. It was common knowledge that you could not succeed without such hacks. There are many ways to fix the problem, but all of them impose some cost either on universities or publishers. Rather than fix the problems, people in other disciplines just criticize psychology when in fact these same problems exist in all the social sciences and similar pressures apply in the physical sciences as well.

[@psicologiajoseh]

That's brutal. Thanks for sharing it.

[@myron1850]

thanks for bringing this up

[@cagribaba4464]

@Psicólogo Miguel Cisneros you sound like a snake oil salesman.

[@Zionswasd]

Well, its not just undervaluing like it is in physical sciences; its mostly just that we could never hope to use the scientific method w social sciences, because of how infinitely complex sociology and psychology are, its impossible to even know let alone control all the variables at play in order to get pure and actually valid scientific data. No offense but you got a phd in a pseudoscience, a religion, essentially.

[@RenegadeContext]

@Zionswasd. It's not a psuedoscience it's a soft science there's a huge difference. There are measurable aspects of it but there are also shifting variables. When psychology is taken seriously by the public and by practitioners (many people who get into it are nut jobs that do it no favours) things will change. There has to be room for error in the same way the shifting variables of meteorology are taken into account

[@mikewilliams6025]

Psychologists and psychology publications, by and large, have not relied on the scientific method. Think of how many policies, drug treatments, patient care, and cultural ethics have been built on "findings" in modern psychology. That should trouble all of us.

[@agathachris9722]

Exactly

[@paolotenis20]

unfortunately yes, we are a product of the medical operational model which creating psychiatry as a means of explaining what at the time couldn't be explained organically, hell even Freud wanted to return to neurology asap. it's time for psychology to either turn back, or start again with something else, to find a principle and strong theories like others real sciences, then we might not even need to have 30 types of psychologies.

[@crimsonhermit]

pure pseudoscience. so sad that the world is buying into it. certain to be a complete joke in the future. much like bloodletting or some other bizarre archaic medical treatment.

[@Zionswasd]

​@@paolotenis20 There's just no way to study psychology with the scientific method, its far far far too complex; being a mix of sociology (another pseudoscience) and neurology (which we dont have super advanced knowledge on).

[@daneflinn3022]

@@Zionswasd I wouldn't call them pseudoscience, but social sciences. Sociology, unlike Psychology doesn't pretend it's anything but a social science.

[@choyou3932]

Thank you for putting this to light. I really detest the idea that i might have to work with researchers who fronts these types of work as a biologist.

[@aktchungrabanio6467]

@Psicólogo Miguel Cisneros lmao that's even worse!!

[@ImplodingChicken]

As usual, a lucid presentation of an interesting topic! I have actually done a deep dive into the replication crisis in the past, so I had a much more in-the-weeds view on it, and it was interesting seeing it presented from this bird's eye view.

[@PhilosophicalQuestions]

Did you dive into ways of solving the problems (e.g. openly sharing data and code, pre-registration, etc.)? How hopeful are you about the future of (social) psych research?

[@mustang8206]

Thanks Imploding Chicken

[@BitOfPixieDust007]

Science and kittens, I'm sold ;) Now I'm eagerly awaiting the sequel concerning the explicit debate. You guys are building some cliffhangers over here!

[@stanleyklein524]

Psychological Science (which I have published in) is NOT THE (or, really, A) top journal in psychology. It is simply a good journal.

[@ConanDuke]

*The Replication Crisis in Science Generally. Strictly speaking, science is not actually epistemology or ontologically possible. See: 'Logical Positivism', Popper, Kuhn, Sasz, Wittgenstein, Watts, McKenna, et al.

[@anthonykennedy5324]

Belief is what we all have in common . Indeed , we cannot NOT believe.Knowledge ( evidence-based belief) is rare and conditional. Hard work.

[@lawrencefitzgerald]

not unique to psychology, same for biology and medicine and other disciplines

[@1davidjose]

The problem usually arises when sampling is a non-probability method

[@c0ck7aiL]

Great stuff, very clearly explained. I presume this all comes from the fact that it is really hard to remove biases in psychology- working with humans is not like working with computers.

[@esahm373]

The mind is a function of the brain, a biological organ. Once the biological mechanisms facilitating the mind are properly understood, there is no such thing as bias. If Psychology was a proper science, it would further an understanding of these mechanism and remove subjective interpretation and theory from its methodology.

[@Psychedlia98]

“Was” ok kid

[@diefresheomie]

love the scene with the natural scientist at 0:41 was it ur ratinale to point out the other folded accuracy in the natural science, which psychology tries to reach with it's massive statistical stuff?

[@PhilosophicalQuestions]

TBH it was free stock video that we thought is fitting, but I like your point more :D

[@mustang8206]

When you try to be smart about a stock video that has no meaning

[@Interpersonalny]

0:41 "There are now more than a dozen different replication attempts of the power posing effect none of them successful" I once read an article on this topic in Forbes, the flow of which was roughly as follows: 1) Amy J. C. Cuddy announced her amazing study 2) Everyone is happy and her TEDx talk is one of the most watched videos on the channel 3) However, there was a problem with replicating the result 4) Eventually Cuddy replicated her study - but not in the field of hormonal change - nevertheless the "power poses" work. And here is the name of the study: P-Curving a More Comprehensive Body of Research on Postural Feedback Reveals Clear Evidential Value for Power-Posing Effects: Reply to Simmons and Simonsohn (2017) So who is right in the end? Is this a replicable study or not? In your video it was said explicitly "none of them successful", and as you can see there were successful replications.

[@Loreweavver]

I think your analogy breaks down when you start coloring the thermometer.

[@frustratedalien666]

When I first heard of the power pose bs, I cringed and thought "pseudo-science". Then I realized it was based on an actually published paper and thought maybe, since psychology is highly subjective, it might just seem bs to me but not seem like bs to a lot of other people. Then I came across the replication crisis in psychological and economic research several years ago and I realized maybe my first conclusion of bs was not so off the mark after all 😅

[@jpalacio7214]

😊

[@sedat4151]

This is what happens when man nominalizes too much of his reality and neglects the need to remain fully open to the denominalized ‘process of reality’.

[@117johnpar]

Its like taking a sledgehammer to brain surgery, In a less analogous way than one might thing.

[@fartsfartington9019]

Maybe because psychology isnt a science and shouldn't be taken too seriously.

[@jennyfindspennys2044]

When done correctly using the scientific method, psychology is absolutely a science and is so serious and important. Mental health is real and we need good empirically supported treatments

[@jskrelz]

Sort of like quantum mechanics

[@Thetarget1]

Quantum mechanics is extremely well-tested, with thousands and thousands of experiments all agreeing. Many of them you can do in a simple high school lab. It also has the most precisely values in all of science, has a mathematically stringent formulation, and is congruent with all other areas of physics, and has many applications in chemistry, electronics, computer science, technology and so on. So no, not at all like quamtum mechanics.

[@jskrelz]

@@Thetarget1 you familiar with the double slit experiment?

[@TemporalOnline]

I don't care if the study is just to see by eye how many legs trully are in the average ant. All I care about is that the data be true. Also, I hate that the average professor would dismiss such a topic for being too mundane. All he had to day was "ok, go ahead". And his tenure should not be in danger for him to be too "common". All this is wrong. all of it, all incentives all the way down.

[@williamlitsch5506]

Frankly, whoever wrote this episode has no idea what science is and missed the point. If it's not accurate, more than 5% of the time, you are essentially not using a scientific method at all. Scientific methods reduce the need for faith and trust. If they are requiring faith and are not replicateable then it's no better than heresy. Under the definition used in this episode, everything is science, but if everyone is special, no one is special.

[@spidgeb3292]

*cough* *cough* Implicit bias *cough*

[@PhilosophicalQuestions]

You don't mean to imply that such a cornerstone of social psych research could be flawed? 🤔 Think of all the policy implications people drew from it!......

[@Rene-uz3eb]

The question is moot given there hasn't come anything useful out of psychological research yet. I think the various self help books have much more interesting ideas, typically more creative and insightful, to ponder, and don't make a claim towards scientific truth. I don't think psychologists understand statistics anyway.

[@stanleyklein524]

How to make psychology a "science"? Stop with the a-theoretiecal demonstrations, figure out (good luck) what are the fundamental issues/aspects of reality that justify psychology as a discipline, and devote study to understanding psychology as a legitimate field related to, though conceptually independent of, biology.