Statistical FAQs
When doing a Screening analysis via Fit Model, what are the "Orthog Coded" and "Orthog t-Test" columns displayed by Effect Screening from the checkmark menu? How are they used? How are they computed?
The Orthog Coded column shows orthogonalized estimates. These are used in the pareto plot because this plot partitions the sum of the effects and partitioning requires orthogonality. These are computed by premultiplying the column vector of Original estimates by the transformation matrix shown by selecting from the pop-up menu in Version 3, or opening the Transformation to make uncorrelated table in Version 4 and higher, and then multiplying by the standard error of the model. The transformation matrix is the Cholesky root of X`X. These estimates depend on the model order unless the original design is orthogonal.
The "Orthog t-Test" values (formerly labeled "normalized") are constructed to make the parameter estimates independent and identically distributed. These are used in the normal plot, which requires these properties. The p-values associated with these estimates (given in the Prob>|t| column) are equivalent to type I sequential tests. Consequently, if the parameters of the model are correlated, these estimates and their p-values depend on the order of terms in the model. These estimates are computed by dividing the orthogonalized estimates ("Orthog Coded" estimates) by their standard errors.
FAQ # 1234
Last Updated: 2007 Jan 22
Browse JMP FAQs
Support Documents
Contact JMP Support
919.677.8008 (US)
