In article <[EMAIL PROTECTED]>, Rob <[EMAIL PROTECTED]> wrote: >Hi all,
>if I have a number of measures of something (25 but this can change), >each measurement is roughly normally distributed, I know the mean and >std of each of these distributions, I know I can I get a more precise >distribution for what I'm measureing, taking account of each >individual measurement by multiplying them together, but is there a >formula for the mean and std of the final relation? and if so what is >it?! If they are independent, it can be done, without the assumption of normality. If they are dependent, which seems to be the case as you are trying to get a more accurate measure of something, it can be done with normality only if the covariances are known as well. Also, should you be multiplying them together, or taking the geometric (or arithmetic, or other) mean? In the case of the geometric mean, the distribution is messy even for the case of normality and independence. -- This address is for information only. I do not claim that these views are those of the Statistics Department or of Purdue University. Herman Rubin, Department of Statistics, Purdue University [EMAIL PROTECTED] Phone: (765)494-6054 FAX: (765)494-0558 . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
