Reply to thread

Message

<blockquote data-quote="Celebrim" data-source="post: 3821321" data-attributes="member: 4937">I know what selection bias is.&nbsp; I'm not publishing this to a journal to establish some scientific fact.&nbsp; Allowing self-selection may or may not create a bias in the sampling results.&nbsp; A scientific study that allows some degree of self-selection may be considered sufficiently valid if an argument can be put forth that what is being studied is not influenced by any factor that would cause selection bias.&nbsp; For example, it's not unusual to ask for volunteers amongst patients when conducting human trials of a new therapy.&nbsp; In this case, the selection bias is probably for for something like 'desparation', but the participants can usually be considered sufficiently representative of the patient population as a whole.&nbsp; Obviously, we are measuring something less concrete than the effectiveness of a treatment, but again I'm not trying to prove anything.&nbsp; All I'm saying is make a 'gut check'.&nbsp; Do we seem like a reasonable cross section of the D&amp;D playing RPG community?&nbsp; If not, then why not?&nbsp; I'm not trying to demonstrate anything about the future direction of table top roleplaying, because the most obvious selection bias is one I'm admitting.&nbsp; We are the D&amp;D playing RPG community.&nbsp; I'm not saying anything, nor do I think the poll says anything, about the non-D&amp;D playing, non-RPG community.&nbsp; Maybe 4e is exactly what it takes to bring that community into the game and launch D&amp;D into the mainstream, and so who cares what the existing fanbase things.&nbsp; But I do think internet polls say something about the D&amp;D playing RPG community and all the squeels of 'biased sampling' 'just because' don't change that.</blockquote>

[QUOTE="Celebrim, post: 3821321, member: 4937"] I know what selection bias is. I'm not publishing this to a journal to establish some scientific fact. Allowing self-selection may or may not create a bias in the sampling results. A scientific study that allows some degree of self-selection may be considered sufficiently valid if an argument can be put forth that what is being studied is not influenced by any factor that would cause selection bias. For example, it's not unusual to ask for volunteers amongst patients when conducting human trials of a new therapy. In this case, the selection bias is probably for for something like 'desparation', but the participants can usually be considered sufficiently representative of the patient population as a whole. Obviously, we are measuring something less concrete than the effectiveness of a treatment, but again I'm not trying to prove anything. All I'm saying is make a 'gut check'. Do we seem like a reasonable cross section of the D&D playing RPG community? If not, then why not? I'm not trying to demonstrate anything about the future direction of table top roleplaying, because the most obvious selection bias is one I'm admitting. We are the D&D playing RPG community. I'm not saying anything, nor do I think the poll says anything, about the non-D&D playing, non-RPG community. Maybe 4e is exactly what it takes to bring that community into the game and launch D&D into the mainstream, and so who cares what the existing fanbase things. But I do think internet polls say something about the D&D playing RPG community and all the squeels of 'biased sampling' 'just because' don't change that. [/QUOTE]

Verification