Refine
Institute
Has Fulltext
- no (18)
Document Type
- Article (18) (remove)
Keywords
- Empirical process (2)
- Goodness-of-fit test (2)
- Independence test (2)
- Paired sample (2)
- Parametric bootstrap (2)
- likelihood ratio test (2)
- not identically distributed (2)
- Bahadur efficiency (1)
- Bootstrap (1)
- Bootstrapping (1)
- Brownian Pillow (1)
- Categorial variable (1)
- Collective risk model (1)
- Concomitant (1)
- Coverage probability (1)
- Cramér-von-Mises statistic (1)
- Cramér-von-Mises test (1)
- Crámer–von-Mises distance (1)
- Equivalence test (1)
- Functional Delta Method (1)
- Gamma distribution (1)
- Goodness-of-fit tests for uniformity (1)
- Hadamard differentiability (1)
- Hoeffding-Blum-Kiefer-Rosenblatt independence test (1)
- Hotelling’s T² test (1)
- Incomplete data (1)
- Integrated empirical distribution (survival) function (1)
- Kernel density estimator (1)
- Length of confidence intervals (1)
- Marginal homogeneity test (1)
- Multi-sample problem (1)
- Numerical inversion of Laplace transforms (1)
- Pitman efficiency (1)
- Resampling test (1)
- Vapnik–Čhervonenkis class (1)
- Volume of confidence regions (1)
- Wilcoxon tests (1)
- allocation (1)
- asymptotic relative efficiency (1)
- bootstrap (1)
- central symmetry test (1)
- complete block symmetry (1)
- conditional excess distribution (1)
- conditional expectation principle (1)
- confidence interval (1)
- covariance principle (1)
- exchangeability test (1)
- forecast (1)
- functional data (1)
- goodness-of-fit test (1)
- huge dimensional data (1)
- independence test (1)
- multinomial distribution (1)
- multivariate normal distribution (1)
- portfolio risk (1)
- random effects (1)
- random effects meta-regression model (1)
- separable Hilbert space (1)
- uniformly most powerful invariant test (1)
Is part of the Bibliography
- no (18)
The paper deals with the asymptotic behaviour of estimators, statistical tests and confidence intervals for L²-distances to uniformity based on the empirical distribution function, the integrated empirical distribution function and the integrated empirical survival function. Approximations of power functions, confidence intervals for the L²-distances and statistical neighbourhood-of-uniformity validation tests are obtained as main applications. The finite sample behaviour of the procedures is illustrated by a simulation study.
A nonparametric goodness-of-fit test for random variables with values in a separable Hilbert space is investigated. To verify the null hypothesis that the data come from a specific distribution, an integral type test based on a Cramér-von-Mises statistic is suggested. The convergence in distribution of the test statistic under the null hypothesis is proved and the test's consistency is concluded. Moreover, properties under local alternatives are discussed. Applications are given for data of huge but finite dimension and for functional data in infinite dimensional spaces. A general approach enables the treatment of incomplete data. In simulation studies the test competes with alternative proposals.
Hotelling’s T² tests in paired and independent survey samples are compared using the traditional asymptotic efficiency concepts of Hodges–Lehmann, Bahadur and Pitman, as well as through criteria based on the volumes of corresponding confidence regions. Conditions characterizing the superiority of a procedure are given in terms of population canonical correlation type coefficients. Statistical tests for checking these conditions are developed. Test statistics based on the eigenvalues of a symmetrized sample cross-covariance matrix are suggested, as well as test statistics based on sample canonical correlation type coefficients.
On the basis of independent and identically distributed bivariate random vectors, where the components are categorial and continuous variables, respectively, the related concomitants, also called induced order statistic, are considered. The main theoretical result is a functional central limit theorem for the empirical process of the concomitants in a triangular array setting. A natural application is hypothesis testing. An independence test and a two-sample test are investigated in detail. The fairly general setting enables limit results under local alternatives and bootstrap samples. For the comparison with existing tests from the literature simulation studies are conducted. The empirical results obtained confirm the theoretical findings.
We discuss the testing problem of homogeneity of the marginal distributions of a continuous bivariate distribution based on a paired sample with possibly missing components (missing completely at random). Applying the well-known two-sample Crámer–von-Mises distance to the remaining data, we determine the limiting null distribution of our test statistic in this situation. It is seen that a new resampling approach is appropriate for the approximation of the unknown null distribution. We prove that the resulting test asymptotically reaches the significance level and is consistent. Properties of the test under local alternatives are pointed out as well. Simulations investigate the quality of the approximation and the power of the new approach in the finite sample case. As an illustration we apply the test to real data sets.
Suppose we have k samples X₁,₁,…,X₁,ₙ₁,…,Xₖ,₁,…,Xₖ,ₙₖ with different sample sizes ₙ₁,…,ₙₖ and unknown underlying distribution functions F₁,…,Fₖ as observations plus k families of distribution functions {G₁(⋅,ϑ);ϑ∈Θ},…,{Gₖ(⋅,ϑ);ϑ∈Θ}, each indexed by elements ϑ from the same parameter set Θ, we consider the new goodness-of-fit problem whether or not (F₁,…,Fₖ) belongs to the parametric family {(G₁(⋅,ϑ),…,Gₖ(⋅,ϑ));ϑ∈Θ}. New test statistics are presented and a parametric bootstrap procedure for the approximation of the unknown null distributions is discussed. Under regularity assumptions, it is proved that the approximation works asymptotically, and the limiting distributions of the test statistics in the null hypothesis case are determined. Simulation studies investigate the quality of the new approach for small and moderate sample sizes. Applications to real-data sets illustrate how the idea can be used for verifying model assumptions.