Quote:
Originally Posted by Anaraxes
Don't psychologists often develop these kinds of tests backwards? That is, rather than construct questions according to some theory, they give a bunch of sample questions to various people whose personalities they determine by other means (interviews, etc), and then look for correlations in the question set.
|
Yes. But I very much doubt SJ Games and Laws had the budget available to do that kind of systmatic research. Most likely Laws based his writings on his own observations, over a decade or more of GMing and of being a player in campaigns run by other GMs.
That said, the questions definitely need to be very carefully phrased, so as to avoid the quiz taker being swayed towards giving answers that do not truly reveal his preferences. And that's me talking general principles. I haven't taken the quiz yet.
Also, doing a sample questions survey on any given forum, including this one, basically cannot avoid selection bias, as each RPG forum attracts certain kinds of people and repels other kinds of people. I mean, imagine such a sample quesions survey, that you run first on the population of this forum, and then on the population of the FUDGE mailing list, then on the population of the Hero System forums.
About the only survey done that may have avoided selection bias was the huge one run by Wizards of the Coast very long ago, and its results are so crazily neat that I can't avoid suspicion that they were doctered heavily, either by cheating after the fact, or else by carefully phrasing the questions in order to get a super-neat result (which I can't see a way of doing other than first running a small survey, then re-phrasing to get prettier results in a larger survey, and then as the third iteration - at least - run the full-scale survey with your by-now carefully tailored questions). Nobody else has the budgetary resources for that kind of thing, I think. Even Storyteller was a niche, although a large one, when it was big.