Data Transformation as a Means to obtain Reliable Consensus Values for Reference Materials

Abstract
Two different methods for estimating modal values of skewed distributions have been applied to round robin data on reference materials. The first method involves polynomial fitting to a histogram, and the second is confined to data transformation to overcome the problems connected with treatment of data characterized by skewed distributions. The transformation method (Gamma) allows the construction of a confidence interval which is non‐symmetrical with respect to the central value of skewed distributions.

This publication has 6 references indexed in Scilit: