zbMATH — the first resource for mathematics

Bounds for probabilities of small relative errors for empirical saddlepoint and bootstrap tail approximations. (English) Zbl 0927.62017
Summary: To obtain test probabilities based on empirical approximations to the distribution of a Studentized function of a mean, we need the approximations to be accurate with sufficiently high probability. In particular, when these test probabilities are small it is best to consider relative errors. Here we show that in the case of univariate standardized means and in the general case of tests based on smooth functions of means, the empirical approximations have asymptotically small relative errors on sets with probability differing from 1 by an exponentially small quantity and that these error rates hold for moderately large deviations.
In particular, for standardized deviations of order $$n^{1/6}$$, the probabilities approximated are exponentially small with exponents of order $$n^{1/3}$$ and the corresponding relative errors tend to zero on sets whose complements have probabilities of the order of the probabilities being approximated.

MSC:
 62E20 Asymptotic distribution theory in statistics 60F10 Large deviations 60F05 Central limit and other weak theorems 62F03 Parametric hypothesis testing
Full Text:
References:
 [1] Albers, W., Bickel, P. J. and van Zwet, W. R. (1976). Asy mptotic expansions for the power of distribution free tests in the one sample problem. Ann. Statist. 4 108-156. · Zbl 0321.62049 [2] Barndorff-Nielsen, O. (1986). Inference on full or partial parameters based on the standardized signed likelihood ratio. Biometrika 73 307-322. JSTOR: · Zbl 0605.62020 [3] Bickel, P. J. and van Zwet, W. R. (1978). Asy mptotic expansions for the power of distribution free tests in the two sample problem. Ann. Statist. 6 937-1004. · Zbl 0378.62047 [4] Daniels, H. E. and Young, G. A. (1991). Saddlepoint approximation for the studentized mean, with an application to the bootstrap. Biometrika 78 169-179. JSTOR: [5] Davison, A. C. and Hinkley, D. V. (1988). Saddlepoint approximations in resampling methods. Biometrika 75 417-431. JSTOR: · Zbl 0651.62018 [6] DiCiccio, T. J., Field, C. A. and Fraser, D. A. S. (1990). Approximations of marginal tail probabilities and inference for scale parameters. Biometrika 77 77-95. JSTOR: · Zbl 0692.62033 [7] DiCiccio, T. J., Martin, M. A. and Young, G. A. (1992). Fast and accurate approximate double bootstrap confidence intervals. Biometrika 79 285-295. JSTOR: · Zbl 0751.62013 [8] Federer, H. (1969). Geometric Measure Theory. Springer, Berlin. · Zbl 0176.00801 [9] Feuerverger, A. (1989). On the empirical saddlepoint approximation. Biometrika 76 457-464. JSTOR: · Zbl 0674.62019 [10] Hall, P. (1990). On the relative performance of the bootstrap and Edgeworth approximations of a distribution function. J. Multivariate Anal. 35 108-129. · Zbl 0705.62045 [11] Hall, P. (1992). The Bootstrap and Edgeworth Expansion. Springer, New York. · Zbl 0744.62026 [12] Jensen, J. L. (1992). The modified signed log likelihood statistic and saddlepoint approximations. Biometrika 79 693-703. JSTOR: · Zbl 0764.62021 [13] Jing, B. Y., Feuerverger, A. and Robinson, J. (1994). On the bootstrap saddlepoint approximations. Biometrika 81 211-215. JSTOR: · Zbl 0796.62015 [14] Jing, B. Y. and Robinson, J. (1994). Saddlepoint approximations for marginal and conditional probabilities of transformed variables. Ann. Statist. 22 1115-1132. · Zbl 0817.62007 [15] Kokic, P. N. (1988). Limit theorems for U-statistics in finite populations. Ph.D. thesis, Univ. Sy dney. [16] McCullagh, P. (1987). Tensor methods in statistics. Chapman and Hall, London. · Zbl 0732.62003 [17] Petrov, V. V. (1995). Limit Theorems of Probability Theory. 4 Clarendon Press, Oxford. · Zbl 0826.60001 [18] Robinson, J., H öglund, T., Holst, L. and Quine, M. P. (1990). On approximating probabilities for small and large deviations in d. Ann. Probab. 18 727-753. · Zbl 0704.60018 [19] Skovgaard, I. M. (1990). Analy tic Statistical Models. IMS, Hay ward, CA. · Zbl 0755.62003
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. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.