This paper considers the problem of estimating the autoregressive parameter in discretely observed Ornstein–Uhlenbeck processes. Two consistent estimators are proposed: one obtained by maximizing a kernel-based likelihood function, and another by minimizing a Kolmogorov-type distance from independence. After establishing the consistency of these estimators, their finite-sample performance and possible normality in large samples, is investigated by means of extensive simulations. An illustrative example to credit rating is discussed.

We introduce new goodness of fit tests for exponentiality by using a characterization based on normalized spacings. We provide relevant asymptotic theory for these tests and study their efficiency, An empirical power study and comparisons are also provided.

Edgeworth expansions as well as saddle-point methods are used to approximate the distributions of some spacing statistics for small to moderate sample sizes. By comparing with the exact values when available, it is shown that a particular form of Edgeworth expansion produces extremely good results even for fairly small sample sizes. However, this expansion suffers from negative tail probabilities and an accurate approximation without this disadvantage, is shown to be the one based on saddle-point method. Finally, quantiles of some spacing statistics whose exact distributions are not known, are tabulated, making them available in a variety of testing contexts. © 1998 Elsevier Science B.V. All rights reserved.

A general parametric estimation method which makes use of the coverage probabilities or spacings is proposed. Under some regularity conditions, it is shown that such estimators are asymptotically normal. This method generalizes the maximum spacing method of estimation that has been discussed in the literature. Furthermore, it is shown that the maximum spacing estimator is asymptotically most efficient within the subclass of spacings-based estimators under consideration. © 2001 Elsevier Science B.V.

In this article we discuss uniformly most powerful unbiased tests for testing exponentiality against a specific class of two-parameter exponential models with increasing failure rate. We show that the optimal test statistic for this problem admits an alternative representation in terms of a spacings statistic. Using the conditional saddlepoint approximation proposed by Gatto and Jammalamadaka (J. Amer. Statist. Assoc. 94 (1999) 533), we provide highly accurate approximations for the significance values. The test procedure is illustrated with two practical examples from reliability and survival analysis. We also obtain the asymptotic distribution of the test statistic under a sequence of converging alternatives, which allows for the computation of asymptotic relative efficiency. © 2002 Elsevier Science B.V. All rights reserved.

Most of the tractable distributions currently available for modeling circular data are symmetric around a modal direction, prominent among them the von Mises distribution. Here we discuss a method of introducing asymmetry into any such symmetric circular model and develop general classes of non-symmetric circular distributions. In particular, we introduce and study a resulting variation of the classical von Mises distribution, along with a biological application. © 2008 Elsevier B.V. All rights reserved.