deimos3428 wrote:The more data you have the more precise a model you can build, but I'm not sure it's less "valid", it's simply less precise. You have to work with what data you have, and apply common sense, naturally.
sauromatian wrote:Validity is a statistical term, in contrast to reliability [what you're calling 'precision']. You don't have to work with what data you have. You could decide that the data isn't sufficient to be useful, & decline to work with it at all. It's essentially what you've described as the "apply common sense" part.Small sample size, which we unquestionably have, hurts both validity & reliability.
gyg wrote:Surely when you drop a $1000/$2000+ on a Dungeons and Dragons adventure you should be leaving common sense at the door
gyg wrote:(Now I'm going to spend my whole day trying to remember the rest of that party!!)
Deadlord39 wrote:Right-O, Pru! The range is 2-20 IIRC.Borka the Jop sounds terribly familiar, Al. What was the program?
serleran wrote:Could it have been Sorcerer's Island?