An algorithm for fitting Johnson transformations to non-normal data

Authors
Polansky, AM Chou, YM Mason, RL
Citation
Am. Polansky et al., An algorithm for fitting Johnson transformations to non-normal data, J QUAL TECH, 31(3), 1999, pp. 345-350
Citations number
22
Categorie Soggetti
Engineering Management /General
Journal title
JOURNAL OF QUALITY TECHNOLOGY
ISSN journal
00224065 → ACNP
Volume
31
Issue
3
Year of publication
1999
Pages
345 - 350
Database
ISI
SICI code
0022-4065(199907)31:3<345:AAFFJT>2.0.ZU;2-V
Abstract
A FORTRAN program which selects the estimated best-fit Johnson transformati on for a set of nonnormal data is presented. The parameter estimates are co mputed using sample percentiles, and the quality of the fit is assessed usi ng the Shapiro-Wilk test statistic on the transformed data. The estimated t ransformation with the largest p-value is provided along with the transform ed data.