Summary
Let ϕ denote an arbitrary non-parametric unbiased test for a Gaussian shift given by an infinite dimensional parameter space. Then it is shown that the curvature of its power function has a principal component decomposition based on a Hilbert-Schmidt operator. Thus every test has reasonable curvature only for a finite number of orthogonal directions of alternatives. As application the two-sided Kolmogorov-Smirnov goodnessof-fit test is treated. We obtain lower bounds for their local asymptotic relative efficiency. They converge to one as α↓0 for the directionh 0(u)=sign(2u−1) of the gradient of the median test. These results are analogous to earlier results of Hájek and Šidák for one-sided Kolmogorov-Smirnov tests.
Article PDF
Similar content being viewed by others
References
Anderson, T.W., Darling, D.A.: Asymptotic theory of certain “Goodness of fit” criteria based on stochastic processes. Ann. Math. Stat.,23, 193–212 (1952)
Bouleau, N., Hirsch, F.: Dirichlet forms and analysis on Wiener spaces. (De Gruyter Studies in Math. 14) Berlin: 1991
Drees, H., Milbrodt, H.: Components of the two-sided Kolmogorov-Smirnov test in signal detection problems with Gaussian white noise. J. Stat. Planning Inference29, 325–335 (1991)
Drees, H., Milbrodt, H.: The one-sided Kolmogorov-Smirnov test in signal detection problems with Gaussian white noise. Statistica Neerlandica48, 103–116 (1994)
Durbin, J., Knott, M.: Components of Cramér-von Mises statistics I. J. Roy. Stat. Soc. Ser. B34, 290–307 (1972)
Goodman, V.: Quasi-differentiable functions on Banach spaces. Proc. Am. Math. Soc.30, 367–370 (1971)
Gross, L.: Potential theory on Hilbert space. J. Funct. Anal.1, 123–181 (1967)
Hájek, J., Sidák, Z.: Theory of rank tests. New York: Academic Press 1967
Ibragimov, I.A., Khas'minskii, R.Z.: Statistical Estimation. Asymptotic Theory. Berlin: Springer 1981
Ibragimov, I. A., Khas'minskii, R. Z.: Asymptotically normal families and efficient estimation. Ann. Stat.19, 1681–1724 (1991)
Inglot, T., Jurlewicz, T., Ledwina, T.: On Neyman-type smooth tests of fit. Stat.21, 549–568 (1990)
Janssen, A., Milbrodt, H.: Rényi type goodness of fit tests with adjusted principal direction of alternatives. Scand. J. Stat.20, 177–194 (1993)
Koshevnik, Yu. A., Levit, B. Ya.: On a nonparametric analogue of the information matrix. Theory Probab. Appl21, 738–753 (1976)
Le Cam, L.: On a theorem of J. Hájek, Contributions to statistics. In: J. Jurecková (ed.) Hájek Memorial Volume. pp. 119–135. Dordrecht: R. Reidel 1979
Le Cam, L.: Asymptotic methods in statistical decision theory. Springer Series in Statistics. New York: Springer 1986
Milbrodt, H., Strasser, H.: On the asymptotic power of the two-sided Kolmogorov-Smirnov test. J. Stat. Planning Inference26, 1–23 (1990)
Neuhaus, G.: Asymptotic power properties of the Cramér-von Mises test under contiguous alternatives. J. Multivariate Anal.6, 95–110 (1976)
Pfanzagl, J., Wefelmeyer, W.: Contributions to a general asymptotic statistical theory. (Leet. Notes Stat. 13) New York: Springer
Pietsch, A.: Operator ideals. Amsterdam: North-Holland, 1980
Pietsch, A.: Nuclear locally convex spaces. Ergebnisse der Mathem. und ihrer Grenzgeb.66, Berlin: Springer 1982
Shorack, G.R., Wellner, J.A.: Empirical Processes with applications to statistics. New York: Wiley 1986
Strasser, H.: Tangent vectors for models with independent but not identically distributed observation. Stat. Decisions7, 127–152 (1984)
Strasser, H.: Mathematical theory of statistics. Berlin: De Gruyter 1985
Strasser, H.: Global extrapolations of efficiency. Stat. Decisions8, 11–26 (1990)