16 Jun 2018


Lara-Murphy Show, Potpourri 30 Comments

==> I’m somewhat embarrassed to say that I don’t think I was very familiar with the apparently famous “Stanford prison experiment,” but my ignorance has been validated because the whole thing was a fraud. (In contrast, I always filed away Milgram’s obedience experiment as my go-to scientific confirmation of whatever political science point I wanted to make in an argument…)

==> Speaking of conformity: all the cool kids are listening to the latest episode of the Lara-Murphy Show, where we warn about the dangers of letting policy loans ride without paying them down.

30 Responses to “Potpourri”

  1. Bitter Clinger says:

    Speaking of conformity my go to confirmation is Solomon Asch’s conformity experiments. The best rendition is from Allan Funk’s What do you say to a Naked Lady (1970) from 36:17 to 39:00. https://www.youtube.com/watch?v=gzb-3cyK9mM&t=2365s What a hoot! Someplace (I can’t find it right now) I read interviews with the subjects describing their thoughts and emotions during the experiment. Talk about PTSD.

  2. Josiah says:

    The irony is that if you take the Stanford Prison Experiement at face value, it’s conclusions were inconsist with those of the Milgram Experiment

    • Bob Murphy says:

      How do you mean? The guards didn’t listen to (what was reported as) the instructions of the people running the experiment?

      • Josiah says:

        In Milgram, the test subjects generally didn’t want to administer the shocks, but they did it because that was what they were told to do. It’s a story of how authority makes good people do bad things.

        In the traditional telling of the Stanford Prison experiment, by contrast, people quickly and eagerly began to act like cruel maniacs just because there was no one telling them not to. It was as if the Milgram experimenters has told the subjects “it’s up to you to decide” and then people started handing out lethal shocks.

        • Harold says:

          The article says the prison experiment was first published in New York Tines magazine, side stepping peer review. Always a red flag.

          I don’t think Milgram and Stanford are not consistent with each other, but attempt to demonstrate different things. It is quite possible that individuals could comply with orders against their better judgement because an authority figure told them to, but also possible that people could change their behavior according to what group they were assigned to according to their expectations. We do know that people do behave in horrible ways in some circumstances.

          This article suggests that far from measuring different things they were measuring the same thing – people comply with authority figures to do bad things. The guards did what they thought the experimenters wanted them to do.

          • Andrew says:

            I agree with you that this new information makes the Stanford prison experiment consistent with the Milgram experiment and I agree with Josiah that the original claims that were reported as results of the Stanford prison experiment were inconsistent with Milgram.

        • Bob Murphy says:

          Ah, right, good point Josiah.

        • RPLong says:

          The results don’t contradict each other, they study different things. Milgram was testing the impact of being placed under authority. Zimbardo was demonstrating (NB: the SPE was not a test, it was a demonstration) that people in positions of authority will tend to abuse their power.

          • guest says:

            For what it’s worth, I think there’s more than authority at work.

            Tell people to shock the guy in the next room until he dies, and see what happens.

            Were I in that situation, I would expect that proper precautions had been put in place; No presumption of sheeple-ness required.

            Just one more area where psychologists think they know something, but they don’t.

            (They seem to often assume subjects are interpreting questions/situations a certain way.

            (For example, the “correct” answer to “Have you ever thought about stealing X, Y, or Z?” is “No”. But maybe *you* answer “Yes” because *everybody* has at least *thought about* stealing X, Y, or Z.”)

            • RPLong says:

              Were I in that situation, I would expect that proper precautions had been put in place; No presumption of sheeple-ness required.

              So, in other words, you would have assumed that the authorities had taken the necessary precautions and that you didn’t have to take individual responsibility for doing the due diligence yourself?

              Sounds like Milgram’s point. How does yours differ?

              • guest says:

                I sit in chairs and walk into buildings all the time without taking individual responsibility for doing the due diligence myself, as far as safety is concerned.

                That aspect – of trade-offs of risks – has nothing to do with authority.

              • Tel says:

                It’s fundamentally difficult to outsource trust in a division of labour situation. It’s easy to outsource simple work to another person when I can quickly check their work, but if I don’t even know enough to know what to check then I am better off doing the work myself to avoid getting ripped off (even if I struggle with that work).

                This problem becomes worse as division of labour becomes more fine grained and especially with intellectual labour such as software design, or law, or governance. These types of activity are complex and difficult to properly check, so there’s always a trust component and an incentive to either cut corners or take advantage of the situation. That said, division of labour is valuable so we don’t want to abandon it.

                This where the whole “skin in the game” concept comes into play. There’s a classic experiment where smoke comes into the room and everyone else is in on the trick, and they pretend not to notice while one person sees and smells the smoke and tries to get the others to pay attention. Because you know if there’s a real fire then all the others will also die, you tend to think they must know it’s safe because the don’t react. If all the others start quietly moving towards the door to leave then you think about it very differently.

                With Milgrams experiment, there is the principle that the person running the experiment will be jailed if anyone dies, so they have the “skin in the game” therefore would not put themselves at risk by doing anything stupid (and this instinct turns out to be correct, even though the test subject does not know that the electric shocks are fake).

              • RPLong says:

                guest, okay, that’s a pretty unorthodox view, the idea that you can simply assume that people have things under control and that no responsibility falls to you. Not everyone thinks like you. I’ll leave it at that.

  3. RPLong says:

    Maybe the Stanford Prison Experiment’s treatment in psychology textbooks is wrong, but that doesn’t mean the whole thing was a fraud. Zimbardo states clearly and unequivocally in his book and elsewhere that the SPE was a demonstration, not a true experiment. He wasn’t testing to see if people did X,Y, or Z, he was setting up conditions that he already knew would produce a desired result and documenting what ensued.

    You can criticize it as being scientifically invalid as an experiment, but as a demonstration it served its purpose. All of the underlying psychology underpinning the demonstration can be verified through actual scientific experiments.

    • Craw says:

      Then why have all replication attempts failed?

      • RPLong says:

        It’s a demonstration. There’s nothing to replicate. To my knowledge, there have been many similar demonstrations that have produced similar results, but because they’re not true experiments, this should neither strengthen nor weaken our confidence in the underlying theories themselves.

        • Harold says:

          I agree with RP that these were looking a different things and are not necessarily contradictory. I do have trouble with the lack of replication. A demonstration could be described as an experiment for which you think you already know the outcome. The outcome should be the same every time.

        • Craw says:

          My chemistry teacher demonstrated that sodium burns in water. Then my other chemistry teacher did the same thing with the same result. By contrast, consider a psychic who guesses the top card of a deck. How often will that replicate? Replicability is the reason *why* one demonstrates things.

          • Harold says:

            Craw, do we agree about something?

          • RPLong says:

            Craw, can you cite any kind of information stating that the SPE has not been replicated? In his book, The Lucifer Effect, Zimbardo mentions several other demonstrations replicating his results. Was he lying, or are you over-selling the notion that the SPE has never been replicated? FWIW, Zimbardo cites his sources.

            • Craw says:


              Zimbardo’s interference with publication is telling, no?

              There was also a BBC replication that didn’t match.

              • RPLong says:

                1) This is a news article, not an academic record.

                2) This article doesn’t establish that the experiment has never been replicated.

                I recommend you actually read Zimbardo’s book to gain a better understanding of this. I reiterate: the SPE was not a test, it was a demo.

              • Harold says:

                One reason there may have been no replication is the experiment would not be allowed in the same way. This makes it difficult to replicate.

    • Harold says:

      The BBC experiment is interesting. They proposed that the original study was more like Milgram – in fact Zombardi told the guards to behave in an authoritative manner and that is what happened. The BBC trial did not tell the guards how to behave. That they did not get the same result as Zombardi is expected.

      Guards had better food and sleeping quarters than the prisoners. The guards were unhappy with their arbitrary privilege and never identified as a group. The prisoners identified as a group and worked together to undermine the guards. The guards were suffering high levels of stress as measured by cortisol. Eventually they decided to join up and work together as a commune. That was OK except a couple of ex prisoners attempted a coup to set up much harsher conditions than the original. They demanded military style uniforms and intended to show how it should have been done. The experiment was stopped before this situation could resolve as it was thought the participants may suffer harm. The situation may have ended up with tyranny, war or just more bickering.


  4. Craw says:

    Milgram’s heirs should challenge Twitter’s patents.

  5. Keshav Srinivasan says:

    Bob, out of curiosity, was I the inspiration for this policy loan episode, or was the timing just a coincidence? In any case, the episode didn’t really answer why it’s a bad idea to never repay a policy loan, other than the fact that it’s not a free lunch in that it reduces your cash value or death benefit. But what’s wrong with that?

    Also, on an unrelated note, why can’t you “be your own banker” with a bank account? That is, save up money in your bank account, and then when you want to buy something withdraw money from your bank account to buy it, then “pay yourself back” by making monthly deposits into your bank account. How is that inferior to using whole life and policy loans?

  6. A Country Farmer says:

    Zimbardo has responded: https://drive.google.com/file/d/1qlEcNiK8CBkJOU1YMyz_OSWIIAq6sMXl/view

    I find it strange how quickly people jump to words like “lie” or “fraud” .

  7. Bitter Clinger says:

    I think you guys miss the point. Milgram has nothing to do with the prison experiments. Your explanation of the Milgram’s experiments came later and doesn’t reflect what he was trying to show. In the spirit of the Ghostbusters’s scene https://www.youtube.com/watch?v=5ohlA__xABw I would suggest you consider if Milgram had designed his experiment so that at base level the subject was able to remember five out of the ten. Then you gave him a shock and after he stopped yelling, on the next series he remembered six out of ten. At that point you gave him an even bigger jolt and after he calmed down he remembered seven out of ten. Who in their right mind would not give him the whole 450 volts to see if he would remember ten out of ten? What Milgram showed is universal to the human condition, there are people who say the Obama’s $800 Billion stimulus wasn’t enough that if it had been $1.6 Trillion it would have worked. If only we paid our teachers more, education would work. All you have to do is take guns away and violence will stop. All you have to do is more of what has not worked in the past and it will be sure to work in the future. In industry we see this is a problem and it has been addressed. The most common technique is called “Crew Resource Management” but lots of people teach similar techniques with different acronyms. The military and airline pilots have adopted CRM enthusiastically. ExxonMobile after the Exxon Valdez disaster embraced this method and for the 90’s and 2000’s, at least, was the standard by which all others were judged. (I have been retired for almost ten years so I don’t have any idea what is happening today) The point is that people must be trained and educated to speak up when they see things not working or someone making a mistake. I have trained my wife, my children, and my grandchildren in CRM. So as not to be accused of being a “ Good Example of People Not Believing Their Own Rhetoric” I was forced to write this to explain why you are wrong. I find this very difficult, I would much rather keep my mouth shut and crank up the voltage to fry you like bacon. Wikipedia has an article that explains CRM, correct even though not the way I was taught. John Nance has a book, If Hospital Could Fly where he is promoting CRM in the healthcare field.

    • Harold says:

      ” I would suggest you consider if Milgram had designed his experiment so that at base level the subject was able to remember five out of the ten…”

      I don’t get your point here. That was not how Milgram designed his experiment. You are describing a hypothetical experiment that is different form the one performed.

      Milgram polled final year psych students, colleagues, 40 psychiatrists from med school and others. All thought that only a few would give the maximum shock. He prepared 4 levels of prompt to get them to continue. The result was lots of people did.

      I found your cited article on CRM very useful and informative, so thanks for that, but it supports the interpretations of Milgram given above. People will obey authority unless specifically trained in ways to cooperate and they often find this training quite difficult.

      It would be a very interesting test to see how people trained in CRM behaved in the Milgram test. The problem for such an individual is that however assertive and clear they are, they will only get back the fixed form responses dictated by the experiment. Their training is designed to elicit cooperative and productive response from the authority figure (pilot captain in the original), but this is specifically prevented fro occuring in the experiment.

Leave a Reply