tagler wrote: > Hello-- > This discussion has reinforced my belief that the classical statistical model > is relied on way too much for psychological data analysis! For one thing, we > almost never mee the assumptions of the model (e.g., random sampling, normal > populations, homogeneity of variance). And, just as importantly, it is too > damn confusing! (it's no wonder students dislike statistics).
These are pretty tired old criticisms. First of all, the main concern is with the normality of the *sampling distributions of means*, not of the underlying populations, and the former is guaranteed to us by the Central Limit Theorem, not by the empirical propoerties of a particular population. Second of all, there is a load of statistical research showing the modest (or even moderate) violations of underlying normality (when it actually is an assumption) and of homogeneity of variance matter relatively little to the resulting probabilities, as long as the populations aren't wildly different in their shapes (i.e., skewed oppostie ways), and the sample sizes are approximately equal. The big problems seem to come when one has large hererogeneity of variance (or wildly different shapes) AND different sample sizes. Even then, there are a whack of relatively easy-to-apply correction procedures, most of which are automatically spit out by most stadnard stats packages. Random sampling is another issue, but the easy approximate fix is to make sure that you generalize to a reasonable population (perhaps not "all humans," but rather, "educated Euro-Americans" or somesuch). The much more important matter is random *assignment* to groups. Regards, -- Christopher D. Green Department of Psychology York University Toronto, Ontario, Canada M3J 1P3 phone: 416-736-5115 ext.66164 fax: 416-736-5814 e-mail: [EMAIL PROTECTED] http://www.yorku.ca/christo/ --- You are currently subscribed to tips as: [EMAIL PROTECTED] To unsubscribe send a blank email to [EMAIL PROTECTED]
