On 12 Jun 2003, philipau wrote: > Sorry if my question seem like a newbie question, but I'm dying to > know if it is statistically sound to compare the mean and standard > deviation of different sample sizes from a population.
It is commonplace to compare the means of two (or more) different samples, or the standard deviations (equivalently, the variances) of two (or more) different samples. (What, after all, did you think statistics was about, if not making such comparisons?) Sizes of samples enter the arithmetic, but need not be identical, in either case. "To compare the mean and standard deviation of different samples", all at one fell swoop as it were, is not commonplace; OTOH, that may not have been what you intended to ask. Clarify, please. > I know all about the central limit theorum [sic; the word is > "theorem"], and how the average sample mean will be normally > distributed, but I don't know if it is useful to compare parameters > of different samples "All about the central limit theorem" is rather a lot: did you mean that? I take it you have not studied statistical inference, and do not know about t-tests, analysis of variance, etc. Perhaps you should acquire (borrow from a nearby library) one or more textbooks on elementary statistical inference, and study them. > that I don't know if they are independent or not Since "indepedence" in this context has intimately to do with how the data were collected (and not much to do with anything else), how can you not know? (Unless, of course, you're using second-hand data which have been inadequately described: in which case either you can get better description from those who collected the data, or you can make enough assumptions to permit an analysis to be performed.) HTH. ----------------------------------------------------------------------- Donald F. Burrill [EMAIL PROTECTED] 56 Sebbins Pond Drive, Bedford, NH 03110 (603) 626-0816 . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
