Wednesday, 28 January 2009

Dubious research says little about videogames

"Video game conditioning spills over into real life", New Scientist says. I can't say I'm convinced.

My main objection is that the conditioning they're talking about had nothing to do with the virtual environment in question. In the experiment (abstract here*), subjects were hooked up to a bicycling video game. Apparently they were informed that it was to test a new sports drink delivery system, so I would imagine the scenario is one involving exercise bikes which control the speed at which the player travels. (Details would be appreciated if anyone can read the article.)

Whenever their own team passed by, the players were given a mouthful of sweet drink via the hose; whenever the opponents overtook them, they were given a mouthful of salty tea. Both teams had easily distinguishable logos on the back of their shirts. The researchers then called the participants back in for an fMRI session, and looked at how they responded to the presence of the opposing team's logo. It turns out, they didn't like sitting near it (on a towel) in the waiting room. The response to the logo was correlated with activity in some brain regions on the fMRI.

They conclude that "stimulus in the virtual environment can acquire motivational properties that persist and modify behavior in the real world". Well, I call bullshit. The main stimulus here wasn't in the virtual environment. They were being fed unpleasant drinks whenever the opposing logo was on screen, and pleasant drinks whenever their own team's logo appeared. That's more than enough to condition the subjects against the opposing logo.

Imagine a "Clockwork Orange" scenario where you were tied to a chair watching "Jaws" and I pinched you in the neck every time Roy Schneider spoke, and gave you delicious cake whenever the shark killed someone. Would I be able to reasonably conclude that the movie had conditioned you to hate Schneider's voice and love seeing people get mauled?

Let's speculate. What if they gave the players a positive taste when they were overtaken by opponents, and a negative one when they were overtaken by their own team? Without altering the "virtual world", I'd be willing to bet they'd get entirely the opposite response to logos. This would quite easily invalidate the idea that "stimulus in the virtual environment" was responsible. This test, while obvious, was not performed.

Furthermore, they could've just as easily sat the subjects down in front of the game, sans straws, and performed the same experiment. Failure in a game is often accompanied by a negative visual and audio stimulus, however even this is probably not necessary for conditioning, as being passed by the "other team" is supposed to illicit a negative response in itself (you are losing). This would be much closer to an actual video game experience**, and any conditioning shown would purely be the "video game conditioning" the New Scientist article speaks of.

The act of conditioning their subjects with drinks is so unnecessary to the experiment's goals, and puts such a massive hole in the research, that I have to wonder what the hell they were thinking.

They find an interesting correlation between the subject's self-perceived immersion in the game and their susceptibility to the conditioning, which is much more valid a conclusion in my opinion. I'm curious as to whether one's involvement in a virtual world makes one necessarily vulnerable to being conditioned, in comparison to a neutral situation, or watching TV, or reading a book. Perhaps one's degree of involvement in any fiction- or non-fiction media increases susceptiblity to conditioning. However that research simply has not been done here.

A better title for the study would be "real-life conditioning spills into real life". Sadly, that's not nearly as headline-grabbing.

*New Scientist provide the article's DOI but I couldn't find a way to actually use it on the journal website, which was irritating. It's The Journal of Neuroscience, January 28, 2009, 29(4):1046-1051.

**I don't know about you, but I don't chug salt every time Mario pops his clogs


innate said...

Someone on Slashdot dug up the article link (PDF):

I have the same objections to the study (at least its writeup on, and was about to post them when I came across your comment. It seems they didn't control for _anything_, so really they've just confirmed that conditioning works.

Anonymous said...

Unfortunately, the article isn't included in my institutional subscription. I'll trawl the Slashdot comments to see if they clarify any of my uncertainties or address any of my concerns, though. Thanks!