One of the most famous psychological experiments in history has probably turned out to be a fake.

This is a science fiction starring Stuart Ritchie newsletter for subscribers. I. If you’d like to receive this straight to your inbox every week, sign up here.

Psychological research hasn’t had a good decade. Not only have modern attempts to replicate old research failed miserably in many cases, some of the most famous research to be found in every undergraduate textbook seems to have never been done—or at least never done. as described – at the first place.

Here’s another one.

I remember as a college student learning about an amazing experiment done by David Rosenhan, a professor of psychology at Stanford University, in the early 1970s. You’ve probably heard about it too. Rosenhan sent eight perfectly healthy people, including himself, to psychiatric hospitals across the United States. When they arrived, they were ordered to tell the hospital staff that they had heard voices (in particular, they said they heard the words “empty”, “hollow”, and “deaf”). They were then told to always behave normally.

Rosenhan described how all but one of these “pseudo-patients” were quickly diagnosed with schizophrenia, a very serious diagnosis. Although they had no other symptoms, they were kept in hospitals for up to 52 days. They often felt overwhelmed and powerless, trapped in this bleak environment with no clear idea of ​​when they would be released.

Then came the second part. Apparently, the hospital staff “heard” about Rosenhan’s experiment and “doubted that such a mistake could happen in their hospital.” Rosenhan told them that he would send dummy patients to their hospital in a few months. The staff signaled to many incoming patients that their hospital was a fake, but (big revelation!) Rosenhan didn’t actually send fake patients at all.

Rosenhan outlined his findings in a now classic journal article in 1973. The science titled “Being Normal in Crazy Places”. “It is clear,” he concluded, “that in psychiatric institutions we cannot distinguish between the sane and the insane.”

The newspaper came out at a time when psychiatry was under great suspicion. The antipsychiatric movement began in the 1960s with works similar to those of R.D. Lang. Experience policy and Thomas Sas The myth of insanity arguing that society, not the individual, is to blame for what are seen as symptoms of mental illness.

As historian Robert Skull explains in a new article on Rosenhan, the pseudopatient story was a major scandal that prompted the American Psychiatric Association to radically rethink the entire specialty’s approach to diagnosing mental illness. In her book Great hypocrite2019 journalist Suzanne Cahalan notes how the study had a major impact on the “deinstitutionalization” movement, with mental hospitals closed and replaced with alternatives such as “community psychiatric centers” in the US (something similar started in the UK). a decade after Rosenhan’s experiment with Margaret Thatcher’s Care in Society policy).

The problem is, you guessed it, there is now a lot of suspicion about the pseudopatient study and whether or not the whole episode really happened. Cahalan describes how he was intrigued by Rosenhan’s experiment and tried to track down some of the pseudopatients. But in general it didn’t work out. Despite years of intense effort, she managed to recognize only three of them: Rosenhan himself, one of his students, and another, which we will talk about a little later.

Cahalan shared the extensive notes and reports she had collected with Skull for his new analysis. He reports some serious problems with the experiment:

• Notes on Rosenhan’s own time as a pseudo-patient are completely inconsistent with the idea of ​​”claiming that you heard voices and then behaved normally.” In fact, Rosenhan behaved very dramatically in the hospital – twitching, grimacing, talking very slowly and generally pretending to be in extreme distress. Perhaps keeping him for treatment was not so much an indictment against the hospital.
• One of the pseudopatients that Kahalana was able to identify was in fact ninth Subjects: He reported a positive experience with the hospital, which may explain why he was not included in the final study report – he was mentioned in an odd footnote.
• Curiously, although this pseudo-patient was not included in the editorial, some details of his case were included – although, of course, in support of the overall story.
• To make matters worse, there is an early version of the article that includes this last patient in its entirety, but in which the numbers are the same as in the published version. The science. The numbers (for example, the number of contacts of patients with doctors) cannot be the same for nine and eight patients, so this is an error in at least one of the versions and, in general, is rather suspicious. Scull concludes that the figures were “clearly falsified”.

Oh. Worst of all, none of the other patients could be located — even though Rosenhan made sweeping references that narrowed their identities, such as calling someone a “famous abstract artist.” Scull concludes that “pseudo-patients disappeared or, more likely, never existed except in Rosenhan’s imagination.”

Looking at the study through this lens, some of the details are indeed questionable. And because those details are so sparse – Rosenhan never tells us which hospitals were involved, for example – it casts doubt on the whole story. All of this seems too perfect to be true.

The thing is, even if the “experiment” went exactly as described, it will not be grounds for changing policy or even changing your mind about psychiatric hospitals. This is remarkable when you consider that a small collection of anecdotes about which we have to take Rosenhan’s word alone tore the pages of a leading scientific journal and became one of the most celebrated studies of psychology of all time.

And this is the problem with research in ancient psychology: the attitude discourse produced Unpleasant suggested data shockingly disproportionate. Psychologist Joe Hilgard has written about his reluctance to teach his students the “classics” of psychology (old documents from the mid-20th century). Of course, they were full of fun ideas and often creative ideas – they turned a college basement into a prison for learning conformity! Bribing people into saying that they enjoy doing a boring task, only to find that they enjoy it even more! And so on, but were not extremely rigorous and provided little useful evidence. Many of them contain statistical errors or other errors that should reduce our confidence in their results to near zero.

Some of them, like Rosenhan’s study, despite these obvious shortcomings, had serious cultural and medical implications. So much so that when we now go back and point out the flaws in the original study, people cling to the thought, “Well, okay, maybe there were problems with the study, but Rosenhan was right about mental institutions!”

It’s like how people fall into the trap of a prank, prank, or fake story on social media, and when they’re told it’s fake, they say, “Well, I only believed it because it’s true!”. Sorry, but it’s important that something be true: if you just want to stick to your old beliefs, there’s no point in looking for new data in the world. In other words, if you offer a story (or scientific study) to support an opinion, it should weaken your support for that opinion if the story (or research) turns out to be fabricated. .

And in the case of Rosenhan’s study—”one of the most influential social science papers published since World War II,” as Scull puts it—it seems that large parts were actually fabricated. Rosenhan wanted to show that doctors cannot tell the sane from the mentally ill. In fact, he showed something equally disturbing: the scientists could not distinguish fact from falsification.

Other things I’ve written lately

In what could become a series regular “Checking the Evidence of What Parents Say,” I explored the idea that you should sterilize the bottles you give your child even after they’ve been cleaned. It turns out that many different countries recommend completely inconsistent things and provide very little evidence to support their recommendations. Then very good!

This week marks 25 years since the publication of Andrew Wakefield’s infamous MMR autism article, which sparked a massive vaccination campaign in the UK and around the world. I wrote an article about the lessons we learned and haven’t learned since.

Brexit negotiations could allow the UK and the EU to work together on science again. Here is my analysis.

And finally, with all the discussion around the Covid lab leak theory — and some people poignantly claiming to be right based on new reports from US government agencies — I wrote about what we need to consider when making our decision. to similar questions (spoiler: this is proof).

Science Link of the Week

An article was published this week in the (very prestigious!) journal Nature Medicine stating that the artificial sweetener erythritol, used in many different drinks, can increase the risk of a heart attack. It turns out that this is not very good. Nutritionist Nicola Hess took this rather caustically. Thread on Twitter.

Thanks for reading Science fiction. See you next week, but in the meantime feel free to contact us. [email protected]

This is a science fiction starring Stuart Ritchie newsletter for subscribers. I. If you’d like to receive this straight to your inbox every week, sign up here.