Box–Cox transformation

Quick Reference

A transformation to normality suggested by Box and Sir David Cox in 1964. They proposed a family of transformations that might be used to convert a general set of n observations into a set of n independent observations from a normal distribution with constant variance. The transformation involves a parameter λ that can be estimated from the data using the method of maximum likelihood. The transformed observation, y(λ), is related to the original observation, y, by

Subjects: Probability and Statistics — Economics.

Reference entries