Invariance, Nonlinear Models and Asymptotic Tests

Several Asymptotic Tests Proposed in the Literature Are Shown Not to Be Invariant to Changes in Measurement Units Or, More Generally to Various Transformations Which Leave Both the Model and the Null Hypothesis Invariant. the Test Involved Include the Wald Test, a Variant of the Lagrange Multiplier Test, Neyman's 'C' Test, Durbin's Procedure (1970), Hausman-Type Tests and a Number of Tests Suggested by White (1982). for All These Procedures, Simply Changing Measurement Units in a Way That Leaves Both the Form of the Model and the Null Hypothesis Invariant Can Lead to Vastly Different Answers. This Problem Is Illustrated by Considering Regression Models with Box-Cox Transformations on the Variables. We Observe, in Particular, That Various Consistent Estimators of the Information Matrix Lead to Test Procedures with Different Invariance Properties. We Then Establish General Sufficient Conditions Which Ensure That Neyman's Test Is Invariant to Transformations Which Leave Invariant the Form of the Model Further, We Give Conditions Under Which a Generalized 'C' Test Applicable to General Restrictions, Is Invariant to the Algebraic Formulation of the Restrictions. in Many Practical Cases Where Wald-Type Tests Lack Invariance, We Find That a Modification of the 'C' Test Is Invariant and Hardly More Costly to Compute Than Wald Tests. This Computational Simplicity Stands in Contrast with Other Invariant Tests Such As the Likelihood Ratio Test. We Conclude That Non-Invariant Asymptotic Tests Should Be Avoided Or Used with Great Care. Further in Many Situations, the Suggested Modification of the 'C' Test Yields an Attractive Substitute to the Wald Test and to Other Invariant Tests.

[1]  Alain Monfort,et al.  A General Framework for Testing a Null Hypothesis in a “Mixed” Form , 1989, Econometric Theory.

[2]  A. Gallant,et al.  Nonlinear Statistical Models , 1988 .

[3]  Richard J. Smith Alternative Asymptotically Optimal Tests and Their Application to Dynamic Specification , 1987 .

[4]  Peter C. B. Phillips,et al.  ON THE FORMULATION OF WALD TESTS OF NONLINEAR RESTRICTIONS , 1986 .

[5]  Michael R. Veall,et al.  Formulating Wald Tests of Nonlinear Restrictions , 1985 .

[6]  Gail Gong,et al.  Pseudo Maximum Likelihood Estimation: Theory and Applications , 1981 .

[7]  J. Schlesselman,et al.  Power Families: A Note on the Box and Cox Transformation , 1971 .

[8]  James Durbin,et al.  Testing for Serial Correlation in Least-Squares Regression When Some of the Regressors are Lagged Dependent Variables , 1970 .

[9]  Prem S. Puri,et al.  On Optimal Asymptotic Tests of Composite Statistical Hypotheses , 1967 .

[10]  G. S. Rogers Mathematical Statistics: A Decision Theoretic Approach , 1967 .

[11]  S. D. Silvey,et al.  The Lagrangian Multiplier Test , 1959 .

[12]  J. Aitchison,et al.  Maximum-Likelihood Estimation of Parameters Subject to Restraints , 1958 .

[13]  Calyampudi R. Rao Large sample tests of statistical hypotheses concerning several parameters with applications to problems of estimation , 1948, Mathematical Proceedings of the Cambridge Philosophical Society.

[14]  A. Wald Tests of statistical hypotheses concerning several parameters when the number of observations is large , 1943 .

[15]  E. Pitman TESTS OF HYPOTHESES CONCERNING LOCATION AND SCALE PARAMETERS , 1939 .

[16]  H. Hotelling Relations Between Two Sets of Variates , 1936 .

[17]  P. Schmidt,et al.  Alternative forms of the wald test: how long is a piece of string? , 1988 .

[18]  Kenneth J. White,et al.  Obtaining any Wald statistic you want , 1986 .

[19]  R. Engle Wald, likelihood ratio, and Lagrange multiplier tests in econometrics , 1984 .

[20]  Richard J. Smith Alternative Asymptotically Optimal Tests in Econometrics , 1983 .

[21]  A. Ronald Gallant,et al.  On unification of the asymptotic theory of nonlinear econometric models , 1982 .

[22]  W. Rudin Principles of mathematical analysis , 1964 .

[23]  E. L. Lehmann,et al.  Theory of point estimation , 1950 .