Paper Info
Title
On the test for the homogeneity of a parameter matrix with some rows constrained by synchronized order restrictions
Abstract
The tests on the homogeneity of the columns of the coefficient matrix in a multiple multivariate linear regression with some rows of the matrix constrained by synchronized orderings, using the test statistics obtained by replacing the unknown variance-covariance matrix with its estimator in likelihood ratio test statistics, form a family of ad hoc tests. It is shown in this paper that the tests in the family share the same alpha-level critical values and follow the same distributions for computing their p-values. A sufficient condition is established for other tests to enjoy these properties, and to be more powerful. Two such more powerful tests are examined.
Year
DOI
Venue
2012
10.1016/j.jmva.2012.01.006
J. Multivariate Analysis
Keywords
Field
DocType
sufficient condition,coefficient matrix,parameter matrix,multiple multivariate linear regression,likelihood ratio test statistic,test statistic,order restriction,powerful test,family share,unknown variance-covariance matrix,alpha-level critical value,multivariate linear regression,hypothesis testing,likelihood ratio test,convex cone,power function,covariance matrix,hypothesis test,critical value,projection
Econometrics,Coefficient matrix,Homogeneity (statistics),Estimation of covariance matrices,Likelihood-ratio test,Matrix (mathematics),Multivariate statistics,Data matrix (multivariate statistics),Statistics,Mathematics,Scatter matrix
Journal
Volume
ISSN
Citations 
107,
0047-259X
0
PageRank 
References 
Authors
0.34
0
2
Name
Order
Citations
PageRank
Xiaomi Hu131.88
Arijit Banerjee2989.46