The Stanford Prison Experiment was hugely influential. We learned it was a fraud (2018)
vox.com
vox.com
I get the sense that some people really just don't like to talk or think about the banality of evil.
Some of the more damning parts:
>"Of the students tested, 81% accurately figured out the experimenter’s hypothesis (that guards would be aggressive and that prisoners would revolt or comply), and 90% predicted that the guards would be “oppressive, hostile, aggressive, humiliating” (p. 158), thereby supporting the argument that demand characteristics were likely operating in the SPE and that the participants in the SPE would have probably guessed how Zimbardo and his coexperimenters wanted them to behave."
>"Reason 4: The experimenters intervened directly in the experiment, either to give precise instructions, to recall the purposes of the experiment, or to set a general direction. "
>"When he began to get interested in prisons, only 3 months before the beginning of the experiment, Zimbardo already had strong views on the subject. As he would confide later, “my sympathies were heavily with prisoners. I was anti-prisons, anti-corrections, etc.” (Zimbardo, 2009, p. 34)."
so what? Did they get a grade on their participation? I'd bet not. So why would they care what the hypothesis was?
>> The experimenters intervened directly in the experiment, either to give precise instructions, to recall the purposes of the experiment, or to set a general direction.
PZ says they said, "We noticed this morning that you weren't really, you know, lending a hand, and I was wondering if there's anything's wrong… We really want to get you active and involved because the guards have to know that every guard is going to be what we call a “tough guard”… what I mean by tough is [that] you have to be firm, and you have to be in the action… It’s really important… for the workings of the experiment [because] whether or not we can make this thing seem like a prison—which is the aim of the thing—depends largely on the guards’ behavior."
Sorry, but I don't see how this invalidates anything. If some of the guards were just phoning it in and not taking it seriously, they did need a talking-to.
- If you were a subject in an experiment, wouldn't you be curious about the experiment design and the hypothesis being tested?
- If you had a good guess as to the tested hypothesis, could that influence your behavior?
If the experiment was via college students, it is unlikely that anyone thought about the experiment design much, because students. This may be true of any adults in that time and area, maybe, as a (unnoted) selection bias. There was an implicit cultural deference to authority in the 70s that doesnt exist now (the thinking of the counter culture was also very different). Finally, there was a weak understanding of what terminology (like subject) implied or the inherent dangers, much less the ethical perils.
It affects subject behaviour and contaminates the experiment.
In fact what the experiment showed is that people can behave in pathological ways when Phil Zimbardo coaxes them into it.
We also have the real-life Nazi Germany experience, where 1000's of previously-normal people did unspeakable things. And of course the Japanese behavior in China before that. You might call these "natural experiments."
With human social behavior it's just not possible to have a rigorously controlled experiment. Nor can you ever end up with rock-solid conclusions similar to those in physics. Even more recent behavioral economics experiments are full of artificial situations, like $10 "bets" because you can't use serious sums. All you can say about critiques of SPE is "compared to what?"
A lot of those so-call normal people were literal Nazi party members who became such voluntarily. In particular, camps were run by SS - the elite of believers.
Literally nothing about Germany proves your thesis. Germany proves that it is possible to create political system under which you reward cruelty and consequently people become cruel. It shows that when politicians promote violence, violence can happen. It shows that when you punish opposition, opposition will be helpless.
This only applies to Altreich citizens, who lived in Germany as of 1933. But the Holocaust involved a lot of people who weren't Altreich citizens.
Austria was only annexed in March 1938, and yet supplied its share of war criminals to the Nazi effort.
The Trawniki men, who guarded a lot of the concentration camps in the East, were Ukrainians who only came under German rule in summer 1941 and didn't even speak German. I would say that the influence of German propaganda on them was negligible compared to an Altreich citizen.
OP referred to supposed process in which people spontaneously become super cruel. The claim that "it is possible to make people cruel by authorities" is something completely different.
--
My instructions to the guards, as documented by recordings of the guard orientation, were that they could not hit the prisoners but could create feelings of boredom, frustration, fear, and "a sense of powerlessness—that is, we have total power of the situation, and they have none." We did not give any formal or detailed instructions about how to be an effective guard.
Asking a person role-playing a guard in a prison simulation to be "firm" and "in the action" is mild compared to the pressure exerted by actual wardens and superior officers in real-life prison and military settings, where guards failing to participate fully can face disciplinary hearings, demotion, or dismissal. Although the research team asked all guards to be actively involved and firmly in control of the prisoners, it never instructed guards to employ brutality, and it explicitly banned the use of physical force.
--
Dog whistles are similar to shibboleths.
Is your thinking that a prison guard (or analogous) would be told to be kind and forgiving, or not given any direction at all on the “right” way to behave?
Maybe that is true. In which case he would be a terrible scientist rather than a fraudulent one. Even so, that still makes the legacy of the SPE false, and one has to wonder why he never corrected the record in the nearly 50 years between the experiment and its failure to replicate.
Even if this is true, whose fault is it? Is it Zimbardo's? Or is it all the academics and media hacks who ran with their own mistaken understanding of what the experiment was actually testing, without bothering to check their understanding against what Zimbardo actually published? (Not to mention all the supporting information that he went to great trouble to make publicly available, as he describes in his response.)
> one has to wonder why he never corrected the record
Trying to correct mistaken or intentionally false claims in the media is a never-ending game of whack-a-mole. Very few scientists have the time or the wherewithal to do this. Zimbardo published several peer-reviewed papers, as he describes in his response; that's the "record". Anyone who takes media descriptions of any scientific research as accurate is kidding themselves.
> its failure to replicate
Zimbardo's response specifically addresses this point (both in discussing the claimed failure to replicate by a British research team, and in describing other research which has replicated the main findings), so I don't think you can just help yourself to this claim.
Wrong. Merely trying is in-fact incredibly easy to do, but there is no record of Zimbardo taking a single step in that direction. You don't have a case.
You must be joking. In addition to the peer-reviewed papers he published, he gave media interviews, he testified in Congressional hearings, he published an article in the New York Times Magazine, he published a popular book, he made a PBS series, he made all of the study materials publicly available, he was an expert witness in three trials...
It seems to me that the one who doesn't have a case is you.
Because if he did then it would be highly surprising to me that despite having written a popular book it took until a few years ago and a research paper (not a popular book, not media interviews, etc that he gave where he corrected the record) for people to realize that the conclusion they had drawn about the experiment was wrong?
So apparently what he couldn’t achieve through multiple popular media sources was conveyed very rapidly through a research paper?
Really I’m being needlessly kind here. He only changed his tune once called out, and his claims that this was what the experiment was about all along are in direct contrast to his communicated intent from the beginning. What do we call it when you publish something that says A, while knowing A is untrue? Scientific fraud.
How is that supposed to be convincing?
The study itself described what they thought it might be able to demonstrate. The fraud is Zimbardo's. He was pretty clear that he was testing the hypothesis that the guards would become tyrants. He told the guards to become tyrants.
He literally looks like the devil in Rick and Morty. There's some other classic villains I'm sure are a more apt comparison but I can't think of any at the moment. https://rickandmorty.fandom.com/wiki/Lucius_Needful
That could be said about plenty of psychologists and sociologists. I hate that the general public thinks of the social sciences on the same level as other branches of science.
I find it to be thoughtful and thorough, although imbued with the passion that launched the experiment (and instructions) in the first place, which I surmise may be part of the critique you find unanswered?
May I ask which specific critiques you believe survive his response and impugn the study?
> the nearly 50 years between the experiment and its failure to replicate.
Obviously Zimbardo (and, it seems to me, a substantial chorus of reputable sociologists and prison reformers) have a different impression. I think we can safely characterize the response to the replication critique to be:
a) The attempts to replicate the study which have actually attempted to imbue the participants with the cultural and psychological accurements of prison life (including, sometimes, express expectations of toughness and control, which are real and documented aspects of standing orders at many prisons) have in fact successfully replicated the SPE - the NSW is the one Zimbardo cites in his response of course.
b) The popular "failure" in replication by the BBC, which is cited all over the internet every time this comes up (though I presume from your cautious wording, is not part of your critique) is very poorly constructed and has little in common with the parameters of the SPE.
c) Real-world prisons which have been studied at length, including some domestic maximum security or supermax prisons and also some detainment during the "global war on terror", have exhibited very similar outcomes to some of those observed in the SPE.
d) It is not possible in an academic context to perform the SPE anymore, as neither the instructions given nor the consent to prolonged stressful conditions are generally permitted in human experimentation anymore.
I find the SPE to be highly imperfect, but failure to replicate is not among the criticisms which seem to me to be justly leveled against it.
> One of three guards on a shift that day left the area during a prisoner count and wasn't even requiring prisoners to follow orders issued by the other guards, so David Jaffee, acting as the SPE Warden, took this guard aside and asked him to become more active, involved, and "tough" in order to make the experimental setting seem more like a prison. Here are his exact words: > "We noticed this morning that you weren't really, you know, lending a hand, and I was wondering if there's anything's wrong… We really want to get you active and involved because the guards have to know that every guard is going to be what we call a “tough guard”… what I mean by tough is [that] you have to be firm, and you have to be in the action… It’s really important… for the workings of the experiment [because] whether or not we can make this thing seem like a prison—which is the aim of the thing—depends largely on the guards’ behavior."
Okay, so if you tell people the experiment requires they be strict rather than lenient, and correct them when they don't follow this, then they act strict. Sure, this is a more accurate representation of prison, but I thought the point was to argue that this was some sort of inevitable dynamic, not just "technically possible"?
> Asking a person role-playing a guard in a prison simulation to be "firm" and "in the action" is mild compared to the pressure exerted by actual wardens and superior officers in real-life prison and military settings, where guards failing to participate fully can face disciplinary hearings, demotion, or dismissal. Although the research team asked all guards to be actively involved and firmly in control of the prisoners, it never instructed guards to employ brutality, and it explicitly banned the use of physical force.
This sounds like more evidence that the intensity of behavior is proportional to the amount of pressure. Pressuring people harder makes them respond more strongly; I still don't see how this is somehow a validation of his thesis.
> With each passing night, he became more creatively evil in ways that went beyond being a tough guard. Indeed, he later said that he began to think of himself as a “puppeteer” who made prisoners do whatever he chose. In an extreme perversion of his experimentally assigned role, he devised an unthinkable way to humiliate all prisoners on the fifth night of the study...Fortunately, I had earlier decided to terminate the experiment the next morning.
That's kind of a ridiculous defense on his part. His timing somehow coincided exactly with particularly brutal abuse of some of the experimentees ostensibly in his care, but of course he couldn't have been actively monitoring it or else he'd be culpable for not intervening...so he somehow managed to have the idea to pull the plug exactly at the time that saved him the most face but also at the subdued level of urgency for the entire incident to play out? Either he was grossly negligent in taking care of the experimentees, or he was complicit by choosing not to intervene; there's no middle ground here that makes any sense, despite him clearly wanting to pretend there is.
> Blum portrays the case of Doug Korpi, alias Prisoner 8612, as an instance of me being duped into believing that a prisoner was having an emotional breakdown when in fact the prisoner was simply faking a breakdown in order to leave the study early...For reasons I cannot fathom, Korpi’s story has changed several times over the past 47 years: from genuinely losing control of his emotions, to getting out of the study so that he could lead an insurrection and liberate the other prisoners, to faking a breakdown just to get out early and study for an upcoming Graduate Record Exam, to other reflections and memory distortions.
I don't find it hard to believe that someone's memory of something might shift over nearly half a century and them not intentionally lying. If it was actually an emotionally harrowing experience, it would almost be more surprising if their understanding of how they felt evolved over time, and if it wasn't, it's still pretty mundane for people to remember things incorrectly after decades. Given that any one of their explanations would individually make sense (I would certainly not be surprised if I had a mental breakdown in a situation like this, and if I didn't but had to get to an exam and genuinely didn't think that I would be allowed to get to an exam I needed to take without lying, I wouldn't hesitate to lie to get out because that sounds like _real_ imprisonment, not an experiment), either their memories just being faulty like most people's are or them actually having an emotional breakdown seem far more like than, I don't know, predicting that this would be a famous experiment and deciding to try to make a name for themselves in it?
The kind of experiment that would provide convincing evidence would have a pro-prison psychologist trying to rig the experiment by tipping off the guards that they are supposed to show the best side of human nature and treat the prisoners in a kind, reforming way; but it all goes horribly wrong when the guards ignore the push from the experimenter and are brutal anyway.
What do we expect would be the outcome of an experiment where the pro-prison psychologist running it tips off the guards about how they are supposed to behave? Zimbardo's work is rather ambiguous. Perhaps it generates the expectation that the guards will be brutal. Perhaps it generates the expectation that the guards will conform to the clues and hints injected by the psychologist running the experiment. The work suffers from a fundamental wrong way roundness in relation to the public policy controversy.
Not to mention that these guards didn't have real power; they weren't the ones running the experiment. They had a very limited pretend power. If you want to know whether power corrupts, it's better to look at Zimbardo himself.
And on another note, I have run forums with hundreds of thousands of users. Almost every time I have promoted a user to staff they became (literally) Hitler within six months and I had to ban them.
IMO, power definitely corrupts.
> LEVITT: Yeah, he said it was real, too. But a lot of times what I’ve found is that that when I try to do experiments as an economist that work great for psychologists, I cannot get them to work. And I really have come to believe that it’s because the people in the study are so keen on doing what the researcher wants them to do, and they think that the psychologist wants them to behave in one way, and they think the economist wants them to behave in a different way, and so it’s hard to reproduce some of those psychological findings. I’d love to do the prison study, and I’d love to do it in a way that was unbiased. And I just, it’s one thing, I would bet a lot of money that things wouldn’t turn out the way that they did in that old Zimbardo study.
> DUBNER: Well, you know, let me read you, here’s what a couple of the volunteers who played guards back then 40-some years ago. Here’s what they said recently. One said that he was playing a role from the outset trying to create drama to quote, “give the researchers something to work with.” And another guard said, “I didn’t think it was ever meant to go the full two weeks. I think Zimbardo wanted to create a dramatic crescendo and then end it as quickly as possible. I felt that throughout the experiment, he knew what he wanted and tried to shape the experiment by how it was constructed and how it played out to fit the conclusion that he had already worked out. He wanted to be able to say that college students, people from middle-class backgrounds, that people will turn on each other just because they are given a role and given power.”
> LEVITT: So people won’t believe me, I have never heard those quotes, I didn’t know anyone else thought that way, what I said before was just my intuition that that was not human behavior what got revealed in those studies.
To give credit where it's due, Levitt basically made a whole career out of showing the replication problems of psychology experiments.
This was the dark conclusion of the Milgram experiment: https://en.wikipedia.org/wiki/Milgram_experiment
Ironic, given the many documented mistakes made by the Freakonomics team.
https://www.americanscientist.org/article/freakonomics-what-...
This was the point of the experiment. That simply coaching people in banal ways to be cruel caused them to be cruel.
Yes the study participants were just doing what they thought the scientists wanted to see… exactly as many humans do in many human organizations.
Perhaps the guards believed deep down there was no real danger to the prisoners. Exactly as many people assume about civil society and the social safety net.
How many people would do things that are quite evil or disastrous to another human’s life if instructed to do so by a company’s performance management system, at risk of losing their job?
Remember the context in the early 70s was the relatively recent discovery of the Nazi’s unimaginably cruel treatment of people, and the unknown question of what could cause humans to behave that way.
It turns out, just some authoritative instructions.
Worth reading the original PIs 2018 response to recent criticism in full. Well written. https://www.prisonexp.org/response
And even better if the authority tells them "this is for science".
That should have been in the original publication then
Quod grātīs asseritur, grātīs negātur (aka Hitchens' Razor)
Yeah, exactly. This type of real-world experiment is being done inside corporations and other bureaucratic fiefdoms.
I think there's some truth to our order-following, obedient nature but people also seemed to put way too much stock into this particular study. It's just one study.
The article didn't claim that the study's conclusion was false. The article only points out that the study was flawed, and therefore cannot be relied upon to draw any conclusions either way. The article is about replication issues in modern science, not about whether people in uniforms can be mean to people.
Not "flawed", fraud. Which is a claim that the conclusion is false--and that the author knew it and published it anyway. That's what "fraud" means in this context. That's an extremely strong claim, which the article does not actually justify. Zimbardo's response, which others upthread have posted a link to, is, IMO, required reading to evaluate anything this article says.
https://www.psychologytoday.com/us/blog/fulfillment-any-age/...
It's BS though. The Milgram experiments actually showed that the explanations given with the orders mattered a lot. When the orders were given by people dressed like scientists and explained with "this is necessary for science", compliance rates were highest. When these factors were changed, compliance rates dropped. Compliance rates also dropped when the experiment was performed outside of college towns. What it actually demonstrated is that people do check orders against their pre-existing ideological commitments, and that people who trust scientists and believe in the necessity of science may be willing to commit atrocities in the service of science. Those same people, if given orders by Nazis, would refuse and walk away from the situation (presuming they could do so without being shot.).
In other words, the shocking result is wrong and the real result is mundane. People may commit atrocities for causes they believe in, when given those orders by people they trust.
For example, things like general relativity, continental drift, or evolution from the fields of physics, geology, and biology.
Does psychology have anything like that?
I personally found it to be non-obvious.
https://www.theatlantic.com/health/archive/2015/01/rethinkin...
https://www.psychologytoday.com/us/blog/fulfillment-any-age/...
And I have my doubts that any of those 6 actually completey believed that a Yale psychology experiment was allowed to dangerously shock students, or that the student actors were all that believable.
https://en.wikipedia.org/wiki/Instinctive_drift
More generally, psychological concepts that can be tested in animals, particularly those found repeatedly in many different animals, can be pretty robust. It's experiments on humans that are hard, both for ethical reasons and because it's hard to "experiment" on something that's as smart as you are. But it's also specifically human cognitive psychology that makes the headlines.
While Kahneman and Tversky's mistakes have been widely touted by people who don't like them, many parts of prospect theory also replicate:
https://en.m.wikipedia.org/wiki/Fraternal_birth_order_and_ma...
In my field (psychology/neuroimaging): injuries to the hippocampus obliterates the ability to form memories for associations between items/events (e.g. between multiple pieces of information in an episode space, time, items). However, it will not generally impair memory for individual items or features.
Children will often have worse memory than adults, but these differences are attenuated when you prevent adults from using or benefiting from organizational strategies at encoding and retrieval.
Everything is not obvious until you start applying some scientific method.
But in general the psychology knowledge is a lot less certain than that of natural sciences, precisely because the experiments are very difficult to conduct and the results often turn out to be irreproducible.
Also, people are diverse and behave diversly and it is impossible to get a sample that represents humanity for your experiment.
I'll only admit I've cheated a little: that Freud's contribution was a scientific one is not an assertion that is "generally regarded as correct." But Freud believed himself to be a scientist and frankly I agree. It's not like Einstein, Wegener, and Darwin are uncontested figures, but certainly Freud is much more so.
It's ironic that this of happening immediately before we enter the ai fueled post-truth era.
No science required, just an old dude who had an idea he wanted to share. That's what TED should be, not anything authoritative or declarative. That is indeed their slogan "ideas worth spreading", they should really stick to that.