×

On the optimal number of classes in the Pearson goodness-of-fit tests. (English) Zbl 1245.62045

Summary: The asymptotic local power of Pearson chi-squared tests is considered, based on convex mixtures of the null densities with fixed alternative densities when the mixtures tend to the null densities for sample sizes \(n\rightarrow \infty \). This local power is used to compare the tests with fixed partitions \(\mathcal {P}\) of the observation space of small partition sizes \(| \mathcal {P}|\) with the tests whose partitions \(\mathcal {P}=\mathcal {P}_n\) depend on \(n\) and the partition sizes \(| \mathcal {P}_n|\) tend to infinity for \(n\rightarrow \infty\). New conditions are presented under which it is asymptotically optimal to let \(| \mathcal {P}|\) tend to infinity with \(n\) or to keep it fixed, respectively. Similar conditions are presented under which the tests with fixed \(| \mathcal {P}|\) and those with increasing \(| \mathcal {P}_n|\) are asymptotically equivalent.

MSC:

62G10 Nonparametric hypothesis testing
62G20 Asymptotic properties of nonparametric inference
PDFBibTeX XMLCite
Full Text: EuDML Link

References:

[1] Berlinet A., Vajda I.: On asymptotic sufficiency and optimality of quantizations. J. Statist. Plann. Inference 135 (2005), to appear · Zbl 1098.62004 · doi:10.1016/j.jspi.2005.06.005
[2] Drost F. C., Kallenberg W. C. M., Moore D. S., Oosterhoff J.: Power approximations to multinomial tests of fit. J. Amer. Statist. Assoc. 89 (1989), 130-141 · Zbl 0683.62027 · doi:10.2307/2289856
[3] Feller W.: An Introduction to Probability and its Applications, Vol. 2. Second edition. Wiley, New York 1966 · Zbl 0138.10207
[4] Ferguson T. S.: Course in Large Sample Theory. Chapman & Hall, London 1996 · Zbl 0871.62002
[5] Greenwood P. E., Nikulin M. S.: A Guide to Chi-Squared Testing. Wiley, New York 1996 · Zbl 0853.62037
[6] Halmos P.: Measure Theory. Academic Press, New York 1964 · Zbl 0283.28001
[7] Inglot T., Janic-Wróblewska A.: Data-driven chi-square test for uniformity with unequal cells. J. Statist. Comput. Simul. 73 (2003), 545-561 · Zbl 1054.62057 · doi:10.1080/0094965021000060918
[8] Kallenberg W. C. M., Oosterhoff, J., Schriever B. F.: The number of classes in chi-squared goodness-of-fit tests. J. Amer. Statist. Assoc. 80 (1985), 959-968 · Zbl 0582.62037 · doi:10.2307/2288561
[9] Liese F., Vajda I.: Convex Statistical Distances. Teubner Verlag, Leipzig 1987 · Zbl 0656.62004
[10] Mann H. B., Wald A.: On the choice of the number of intervals in the application of the chi-squared test. Ann. Math. Statist. 13 (1942), 306-317 · Zbl 0063.03772 · doi:10.1214/aoms/1177731569
[11] Mayoral A. M., Morales D., Morales, J., Vajda I.: On efficiency of estimation and testing with data quantized to fixed numbers of cells. Metrika 57 (2003), 1-27 · Zbl 1433.62021 · doi:10.1007/s001840100178
[12] Menéndez M. L., Morales D., Pardo, L., Vajda I.: Asymptotic distributions of \(f\)-divergences of hypotetical and observed frequencies in sparse testing schemes. Statist. Neerlandica 8 (1998), 313-328 · Zbl 0937.62021
[13] Menéndez M. L., Morales D., Pardo, L., Vajda I.: Approximations to powers of \(\phi \)-disparity goodness of fit tests. Comm. Statist. - Theory Methods 8 (2001), 313-328 · Zbl 1008.62540 · doi:10.1081/STA-100001562
[14] Morris C.: Central limit theorems for multinomial sums. Ann. Statist. 3 (1975), 165-188 · Zbl 0305.62013 · doi:10.1214/aos/1176343006
[15] Vajda I.: On the \(f\)-divergence and singularity of probability measures. Period. Math. Hungar. 2 (1972), 223-234 · Zbl 0248.62001 · doi:10.1007/BF02018663
[16] Vajda I.: On convergence of information contained in quantized observations. IEEE Trans. Inform. Theory 48 (2002), 2163-2172 · Zbl 1062.94533 · doi:10.1109/TIT.2002.800497
[17] Vajda I.: Asymptotic laws for stochastic disparity statistics. Tatra Mount. Math. Publ. 26 (2003), 269-280 · Zbl 1154.62334
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.