Close, Dyldonics. I don't think there was any relevant demand characteristic in this experiment. However, we have almost certainly have problems with constructs and with cor./caus.
Constructs - 'tricky mental tests', is it? Which ones? Will the results be different if I use entirely different tests which are still 'tricky'?
Cor./Caus. - the big problem. My first immediate thought was "People I know who are gamers are already nerds. They are computer oriented people who used Apple IIes in the 80's and are now working in software 15 years later. They spend their lives PRACTISING multitasking, and the computer games are incidental."
Originally Posted by johnmansley
It could also be that people who are good multi-taskers are more likely to play video games rather than video gamers being more likely to multi-task well.
Excellent point, cumboy! I think formally that would be a form of sample bias, but I'm not 100% sure.
The way around all of this is of course to use a randomised controlled trial.
1) Specify the outcome you expect, for the reasons you expect.
2) Take a big-as-possible group of random people who do NOT play computer games.
3) Test their 1337 skillz.
4) Select half at random.
5) Make that half play video games.
6) After a period of time, re-test them.
7) Have the hax0rz group increased in skill? Are they now m4d with phatazz sk111z?
I sincerely doubt any psychologist breathing is gay enough to attempt said research.
far_beyond_sane - contributing to the moral decay of your children since 1982
"It was some kind of evolutionary glitch, she figured; no different than the other unreasonable side effects of consciousness and emotion, like religion and rap music."