------------------------------------------------------------------------------- help for mvtest (David E. Moore) -------------------------------------------------------------------------------Multivariate F tests --------------------
mvtest varlist1 [/varlist2 [/varlistn]] [, options ]
Description -----------
mvtest tests linear hypotheses about the estimated parameters from the most recently estimated multivariate regression using Wilks' lambda, Pillai's trace and Hotelling-Lawley's trace. An optional transformation matrix to be applied to the dependent variables prior to testing may be supplied.
Options -------
hyp prints the hypothesis sum of squares and cross products matrix.
err prints the error sum of squares and cross products matrix.
m(elements) specifies the transformation matrix for the dependent variables. It uses the same format as an element by element matrix definition (see [R] matrix). The default is an identity matrix.
norm normalizes the transformation matrix.
Remarks -------
mvtest performs multivariate F tests derived from the eigenvalues of the hypothesis and error sums of squares and cross products matrices. The diagonal elements of these matrices correspond to the usual univariate F tests for regression and ANOVA. If you supply a transformation matrix for the dependent variables, the hypothesis and error sums of squares and cross products matrices are transformed accordingly. You can use the same transformation matrix to test several separate hypotheses by separating the list of variables to be tested with a slash '/' character. Variables separated only by spaces denote a set of variables to be tested jointly.
Examples --------
mvreg dog cat rat = foo bar zoo_1 zoo_2 zoo_3
mvtest _cons
mvtest foo / bar, hyp err
mvtest foo / zoo_*, h m(1,1,1 1,-1,0 1,0,-1)
Stored results --------------
mvtest stores in the $S_# macros:
S_1 Wilks' lambda S_2 F associated with Wilks' lambda S_3 df1 for F associated with Wilks' lambda S_4 df2 for F associated with Wilks' lambda S_5 Pillai's trace S_6 F associated with Pillai's S_7 df1 for F associated with Pillai's trace S_8 df2 for F associated with Pillai's trace S_9 Wilks' lambda S_10 F associated with Hotelling's trace S_11 df1 for F associated with Hotelling's trace S_12 df2 for F associated with Hotelling's trace
mvtest stores in the S_# matrices:
S_0 The transformation matrix S_1 The error SS&CP matrix S_2 The hypothesis SS&CP matrix
Author ------
David E. Moore University of Cincinnati email: moordiatemail.uc.edu
Updates -------
Updated to work with version 6 by UCLA Academic Technology Services
Also see --------
[R] mvreg, [R] test
Statistical Methods and Data Analytics
