A Modification of Box-Cox Transformation

Abstract

Data screening is the most important technique to check the nature of the data. One of the methods
to screen the data is the transformation method. There are a lot of transformations but the famous
one is the Box-Cox Transformation. The purpose of this study is to find the alternative way of
transformation by making some modification of the original Box-Cox formula. As we know, the Box-
Cox transformation method only transform the data if and only if the data in a positive values. If the
values in the analysis are negative, the formula of the Box-Cox cannot be used. This problem always
occur when the researcher want to transform the data which the values is negative. In this
methodology section, we will build an alternative method to solve the problem if the data in the study
are negative.