IJPAM: Volume 66, No. 1 (2011)

RISK OF ESTIMATORS FOR POINT SYMMETRY AND
DOUBLE SYMMETRY MODELS FOR SQUARE
CONTINGENCY TABLES

Kouji Tahata$^1$, Takahiro Inoue$^2$, Sadao Tomizawa$^3$
$^{1,2,3}$Department of Information Sciences
Faculty of Science and Technology
Tokyo University of Science
2641, Yamazaki, Noda City, Chiba, 278-8510, JAPAN
$^1$e-mail: kouji_tahata@is.noda.tus.ac.jp
$^3$e-mail: tomizawa@is.noda.tus.ac.jp


Abstract.For contingency tables, Bishop, Fienberg and Holland [1] gave the expected mean squared error (risk) of the maximum likelihood estimators (MLEs) of cell probabilities for the independence model, and Tahata, Miyamoto, Sasajima, and Tomizawa [4] gave the conditional risks of the MLEs for the symmetry and conditional symmetry models.

The present paper gives the risks of those for the point symmetry model and the double symmetry model for square contingency tables. The paper also compares the risks for these models.

Received: October 4, 2010

AMS Subject Classification: 62H17

Key Words and Phrases: double symmetry, mean squared error, point symmetry, risk, square tables

Source: International Journal of Pure and Applied Mathematics
ISSN: 1311-8080
Year: 2011
Volume: 66
Issue: 1