Partnership vs Prison Experiment

The Stanford Prison Experiment (official site, Wikipedia, BBC Recreation)  is notorious in many ways. For those who haven’t heard of it, in 1971, a randomly selected group of 24 males (out of 70) were split into two randomly assigned groups: prisoners and guards. They were then placed into a mock prison situation. Despite agreeing to a 7-14 day experimental run, the experiment was terminated after 6 days. By this stage, 1/3 of the guards were showing sadistic tendencies, 2 prisoners had quit and the abuse that prisoners were suffering included solitary confinement, loss of mattresses, reduced access to toilets (or enforced primitive access).

Lest you think that the researcher controlling it, Professor Philip Zimbardo, terminated this for altruistic reasons, it was in response to the objections of a graduate student who was observing the experiment – who he was dating and went on to marry. Of the fifty people who had observed the experiment, and the deteriorating conditions, Zimbardo claims that only this one observer objected.

I assume that some of you thinking “Surely… someone else said something?” Yeah, I thought that too, when I first read about it. Apparently not.) It’s worth noting that Zimbardo’s prison started out as a more ‘extreme’ prison than usual, with degrading activities forced onto the prisoners fairly early. You can read about these in the site or you can read about the Abu Ghraib incident, which is strikingly similar.

What is worth noting up front is that this research has never been fully successfully replicated, for a number of reasons, and the publication standard was low. The Stanford Prison Experiment stands, however, in many ways as a failure to protect the people in an experiment, even if the actions of the agents in the system was not as random (there are claims Zimbardo engineered large sections of this) or as meaningful (the experiment was very poorly constructed) as it may appear.

The random selection of the participants, 24 out of an original 70 and randomised roles, seems to indicate a situational attribution of behaviour, rather than one that we are born with. Put someone in a position where they have power over somebody else, put enough rules in the way – *bang* you’re potentially recreating the Stanford Prison Experiment. (Paging Dr Milgram… a topic for a later post.)

Ultimately, demanding compliance can place us in difficult positions where we require authoritarianism on the part of those who demand, and compliance from those who must obey. Whatever the Stanford study showed, this arrangement of struct power-divided rules does not allow for a meeting of minds in any kind of partnership.

One of things I dislike the most about some of the seemingly arbitrary things that we sometimes do (or are encouraged to do) in teaching is that any situation that devolves to “because I said so” or “because I’ve been told to” is requiring compliance under the aegis of institutional support, and driven by some legitimising framework. This gets in the way of one of the most useful and constructive relationships that we can form – the partnership between educator and student. Now, I’m not suggesting for a second that most students have the maturity or depth of knowledge to devise and run entire courses but a partnership role allows us to avoid falling into the traps of guard and prisoner. We do have hard limits that we need to adhere to, to make the recognition of education possible in many senses, but building courses that clearly set these limits in a constructive and useful way, rather than a reactive and inauthentic way, pulls us out of the “I told you to do this” and allows us to move into the “why didn’t you do that?”

(We could talk about allowing individuals mobility to reduce their dependency on external validation from their peers, and hence allow us to encourage the pursuit of individual goals and reduce any fighting over favouritism but I’m not well-versed enough in social identity theory yet to give this much flesh.)

I, as the subject matter expert, am trying to assist the student in developing knowledge within a particular set of subjects and any useful associated areas. If I have created something where, in order to understand the work, you need to complete certain readings and assignments, perform certain actions, and do so in a certain timeframe or lose the opportunity to participate – most students will actually do this. On top of the issues of knowledge, we have the other skills that we are trying to transfer: design, time management, ethics, professionalism, communication skills. This is where it gets hard.

Say, for example, I design a course where you need to finish Assignment 2 before we discuss a certain topic in a lecture/tutorial/studio activity. Therefore, you have a reason to finish assignment 2 before some deadline. I can set a deadline that is just before the next activity or I can set it a few days before to give people some digestion time prior to looking at it again. Or I can set an earlier deadline to give people practice at time management. However, if Assignment 2 is work that will not be referred to elsewhere in the course, except for the exam, when should I set the deadline?

The problem we have is that allowing deadlines to run late means no marking or feedback until late – this, of course, drives our education design to bring formative work forward but, once again, this only makes sense if that feedback will be useful earlier on.

So, to briefly recap, setting an arbitrary hand-in time that is purely to make your marking life easier and has no pedagogical driver or no impact on student learning is understandable but, in many ways, potentially an abuse of your position. (I am all too familiar of the realities of staff and resource shortages on when and how we can mark, especially when we start getting told to increase feedback or have all assignments back within time X. But let’s get this straight: formative and summative have different roles and marking loads. We know that we can achieve things with good learning design that far exceed what we can manage with arbitrary action.)

Now let’s look at a more complex issue – late penalties. I have evidence that students change their behaviour when late penalties are fixed on 24 hour barriers. We’ve seen students line up with these and start handing up in response to these new barriers: miss one and you lose even more marks. But have we changed the right behaviour or does this merely lead to a certain form of resignation in the face of arbitrary authority?

Why am I removing marks anyway? If the work is handed in before the time that it’s needed, then, from a knowledge point of view, the aim has been achieved. Which skill am I developing? If you responded with ‘time management’, then providing that we are completely clear on when the work must be handed in to achieve certain requirements AND that we have added an overall factor in the ‘professional’ spectrum of time management, we are probably doing the right thing. If we’re just saying “hand it in on time OR ELSE” then we are conflating issues of knowledge development with issues of compliance and this is where it starts to get murky.

Now it doesn’t have to get murky but it’s completely possible in this zone. You risk ending up academics who won’t accept anything because it’s late (regardless of reason) or students who start acting up (out of defiance) or, potentially worse, students who become completely passive and dependent upon your authority. If self-regulation is supposed to be in play, then we haven’t achieved much by doing this.

Nothing I’ve said should be interpreted as “no deadlines” or “no authority” but what I am saying is that we know what happens when we take a randomly assigned group of people and make one beholden to the other, when there is no really good reason or sense of equality or partnership between them. We’ve seen it time and time again.

Kohn, in “Punishing with Rewards”, makes a number of observations, some good and some bad, including that one of our biggest risks is in the rupturing of relationships by setting up a disparity of power levels, where one person controls and the other person complies or seeks to appease, rather than to achieve the actual objective. It’s an interesting way to look at a very challenging problem, to give us more lines along which to think.

I should finish this by noting, again, that Zimbardo’s experiment was flawed in many ways and deriving significance from the role is hard. It appears, from the UK version, that leadership plays a key part in what happens. It was only when strong leadership started to lead the prison guards down dark paths in the UK recreation that they started to approach what had happened in Stanford. Zimbardo admits that his role in the experiment may have been not been all that sensible in many ways but it may be that his briefing set the scene for what happened. His passive observation as matters deteriorated, with the guards knowing that he was watching, certainly validated their actions. Either way, if it is a fact that one key leader can have so much impact, then that makes what we do even more important – even if it’s occasionally looking at something, thinking about it and saying ‘No, actually, that’s wrong.”