Reblogged: OUCH! Don’t touch that button!

This piece was originally published by the Macarthur Foundation on their Spotlight on Digital Media and Learning blog (original link).

It is a classic study, one that every student of psychology–and anyone who does research on “human subjects”–knows: The Milgram Experiment.

In the 1960s, psychologist Stanley Milgram studied reactions to authority by asking volunteers to administer electric shocks to a subject as part of a research study. The volunteers were told that the study was looking at how punishment would affect the subject’s performance on a test. The volunteers were reluctant, and grew more so as they heard the cries of pain from the subject getting the shocks. But the researchers said they had to continue with the experiment, and the volunteers gave bigger and bigger shocks to the subject–in some cases at or past dangerous levels.

But the experiment was a ruse. There were no subjects and no shocks. The study was really about the volunteers’ willingness to submit to authority: to continue to punish someone despite more and more desperate cries for help because a researcher had told them that they had to do it. And the experiment is notorious because of the psychological damage it did to the volunteers, many of whom were haunted by their willingness to inflict pain on another person, even after they learned of the deception and the true purpose of the experiment.

Fast forward 40 years, to a new study, this time by Mel Slater, of the Catalan Polytechnic University in Barcelona, Spain, and University College London, UK. (The study is reported in Nature’s news site (originally available at: http://www.nature.com/news/2006/061218/full/061218-17.html), with subscription required, but also more briefly here for free.)

The setup is exactly the same as in Milgram’s experiment, only this time the subject getting electric shocks is a character in an immersive virtual environment.

The volunteers knew that they were interacting with a virtual character, and that no one was actually being hurt.

What was surprising about the experiment was that the results were strikingly similar to the original experiment. As Nature news reports: “Almost half of those who could see the woman said afterwards that they had considered withdrawing from the study, and several actually did.” The volunteers became more and more anxious as the study went on:

Measures of stress, such as heart rate and sweatiness of palms, increased.

Slater concluded: “Of course, consciously everybody knows nothing is happening, but some parts of the person’s perceptual system just takes it as real. Some part of the brain doesn’t know about virtual reality.”

What this makes clear is that being in a well-designed virtual world can be a “real” experience, full of hopes and fears, guilt and pleasure. (In fact, the study suggests that even a relatively crude virtual world can have the same effects.) Thus, players don’t necessarily experience games as “less real” than other things they do in the world. In fact, games–particularly games for learning–can even be more than things many children have a chance to do.

In How Computer Games Help Children Learn, for example, I write about the game The Political Machine. In the game, players run a presidential campaign, and have a chance to learn important lessons about the way our political system works.

Now, of course, it might be more effective for learning to participate in a real election for student body president instead of a simulation of an election. But as I argue in the book, working in the real world has disadvantages too. Student elections take longer. They address a narrower range of issues. Not as many students can run a real campaign as can play the game. And perhaps more important:

The knowledge gained in the game (about the electoral college, fundraising, and advertising) applies more directly to conditions that students will encounter outside of school than what happens in most student government elections.

My own work, and the work of Sasha Barab and other researchers funded by the foundation’s Digital Media and Learning initiative, shows that games where players solve simulations of real problems help them solve similar problems in the real world–and it is clear that we can simulate a wide range of problems that are too expensive, dangerous, or difficult for players to solve in the real world.

So in a very real sense, games for learning can be more real than real life.

Coda: Being interested also in the ethical implications of the virtual Milgram experiment, I mentioned the study in passing to the head of my School of Education’s Institutional Review Board–the committee responsible for approving any studies with human subjects. He cringed and said: “Please don’t do that experiment here.”

It doesn’t get much realer than that….