COMPARISONS ON THE ANALYSIS OF POISSON DATA

摘要

Industrial experiments often result in the production of various data ,both continuous data as well as discrete data. Most of the time ,practitioners use a normal analysis for continuous data and a logistic regression analysis for binary data. Taguchi (System of Experimental Design, Kraus International, New York,1986)suggested that one use various signal-to -noise analyses when encountering different types of data. Engineers do not usually encounter Poission data ;however, in this paper a set of Poisson data is presented with three approaches to analyses the data set. Their comparison is also presented.