Analysis of Means Tables Using Mathematical Processors

Abstract

Introduced by Ellis Ott in the 1960s, the analysis of means (abbreviated ANOM) procedure provides a simple graphical technique for testing for differences between the means of several populations. The ANOM charts are easy to use, similar in appearance and interpretation to control charts, and usually give the same results as the more familiar analysis of variance technique. Unfortunately, the ANOM decision limits depend upon tables of constants that are usually only available in printed form or from specialized software programs. In this article we present new formulas for the ANOM constants and then illustrate how commonly available mathematical processors can be used to calculate the ANOM constants. This approach makes the ANOM constants easily accessible, portable, and unrestricted with regard to the choice of significance level, sample size, and number of populations.

Keywords:

• Analysis of means
• Bivariate normal distribution
• Bivariate t distribution