My apologies to the news group - Maurus didn't give a valid email address. You seem to be looking for a ratio, of stdev/average. More accurately, you would like a ratio of a measurement of dispersion divided by a measurement of central tendency (I think those are the official terms for the concepts:).
the coefficient of variation is the est. stdev/average, or stdev/mean, depending on who is talking. If you have a theoretical maximum, then you are discussing a central tendency measurement which has an upper bound. This is not a Normal distribution measurement, for starters, and is most likely a binomial, or hypergeometric, distribution. If binomial, your ratio would be stdev/100%, or some such. In a binomial, the stdev is a function of the expected probability, p, so this ratio would not give you much new information. I once did a study on the melting point of a metal eutectic alloy. The two major components restrict the melting point to a known temperature, and the additional elements (of interest) lower the m. p. slightly. this would be a case where the ratio of stdev/max m. p. might have usefulness. It might have been a hypergeometric dist, I didn't test for that. However, we were sufficiently far into the details that we didn't need to re-phrase the stdev in this manner. Help any? Jay Maurus wrote: > One way to obtain a relative variability index is devide an obsolute one > with its theoretical maximum (sorry, i don't know if it's correctly written > since i'm studying statistics in another language). > > What is the procedure used to find this theoretical maximum? > > Regards. > > .. > .. > ================================================================= > Instructions for joining and leaving this list, remarks about the > problem of INAPPROPRIATE MESSAGES, and archives are available at: > .. http://jse.stat.ncsu.edu/ . > ================================================================= -- Jay Warner Principal Scientist Warner Consulting, Inc. 4444 North Green Bay Road Racine, WI 53404-1216 USA Ph: (262) 634-9100 FAX: (262) 681-1133 email: [EMAIL PROTECTED] web: http://www.a2q.com The A2Q Method (tm) -- What do you want to improve today? . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
