Accurate and efficient construction of bootstrap likelihoods

The simplest construction of bootstrap likelihoods involves two levels of bootstrapping, kernel density estimation, and non-parametric curve-smoothing. We describe more accurate and efficient constructions, based on smoothing at the first level of nested bootstraps and saddlepoint approximation to remove second-level bootstrap variation. Detailed illustrations are given.     

An L 2 Comparison Between the Bootstrap and the Empirical Edgeworth Expansion

The Bootstrap estimate for studentized statistics is more accurate than both the normal approximation and the two-term empirical Edgeworth expansion. In this article, it will be shown that the three-term empirical Edgeworth expansion for studentized statistics compares well with the bootstrap. It is also shown that the three-term Edgeworth expansion is superior to the bootstrap in some cases, using more efficient estimators than sample moments in the Edgeworth expansion, such as using maximum likelihood estimators in the one-parameter exponential family.     

On the relative performance of the block bootstrap for dependent data

In the independent setting, both Efron's bootstrap and “empiricai Edgeworth expansion” (E.E-expansion) give second-order accurate approximations to distributions of standardized and studentized statistics in the smooth function model. As a result, Efron's bootstrap was often regarded as roughly equivalent to the one-term E.E-expansion. However, a more detailed analysis shows that Efron's bootstrap outperforms the E.E-expansion in terms of loss functions by Bhattacharya and Qumsiyeh (1989) and in terms of probabilities for large deviations by Hall (1990) and Jing et a1 (1994). in this paper, we shall study the performances of the block bootstrap and the E.E-expansion for the weakly dependent data. It turns out that similar properties hold:both perform equally well at the center of the distribution but the block bootstrap provides accurate approximations even in the tails of the distributions. The study is focued on the simple case of standardized and studentized sample mean, but the conclusions can be easily extended to the smooth function of multivariate means.     

On the bootstrap and smoothed bootstrap

The standard bootstrap and two commonly used types of smoothed bootstrap are investigated. The saddlepoint approximations are used to evaluate the accuracy of the three bootstrap estimates of the density of a sample mean. The optimal choice for the smoothing parameter is obtained when smoothing is useful in reducing the mean squared error.     

Edgeworth expansion and the bootstrap for stratified sampling without replacement from a finite population

An Edgeworth expansion for a linear combination of stratum means in stratified sampling without replacement from a finite population is derived. The expansion is applied to a bootstrap proposed for this context to show that the bootstrap captures the second-order term of the expansion.     

Level-specific correction for nonparametric likelihoods

The popular empirical likelihood method not only has a convenient chi-square limiting distribution but is also Bartlett correctable, leading to a high-order coverage precision of the resulting confidence regions. Meanwhile, it is one of many nonparametric likelihoods in the Cressie–Read power divergence family. The other likelihoods share many attractive properties but are not Bartlett correctable. In this paper, we develop a new technique to achieve the effect of being Bartlett correctable. Our technique is generally applicable to pivotal quantities with chi-square limiting distributions. Numerical experiments and an example reveal that the method is successful for several important nonparametric likelihoods.     

Monte carlo approximation to edgeworth expansions

Since the 1930s, empirical Edgeworth expansions have been employed to develop techniques for approximate, nonparametric statistical inference. The introduction of bootstrap methods has increased the potential usefulness of Edgeworth approximations. In particular, a recent paper by Lee & Young introduced a novel approach to approximating bootstrap distribution functions, using first an empirical Edgeworth expansion and then a more traditional bootstrap approximation to the remainder. In principle, either direct calculation or computer algebra could be used to compute the Edgeworth component, but both methods would often be difficult to implement in practice, not least because of the sheer algebraic complexity of a general Edgeworth expansion. In the present paper we show that a simple but nonstandard Monte Carlo technique is a competitive alternative. It exploits properties of Edgeworth expansions, in particular their parity and the degrees of their polynomial terms, to develop particularly accurate approximations.     

Do bootstrap confidence procedures behave well uniformly in P?

An important statistical problem is to construct a confidence set for some functional T(P) of some unknown probability distribution P. Typically, this involves approximating the sampling distribution J_n(P) of some pivot based on a sample of size n from P. A bootstrap procedure is to estimate J_n(P) by J_n(&Pcirc;_n), where P?_n is the empirical measure based on a sample of size n from P. Typically, one has that J_n(P) and J_n(P?_n) are close in an appropriate sense. Two questions are addressed in this note. Are J_n(P) and J_n(P?_n) uniformly close as P varies as well? If so, do confidence statements about T(P) possess a corresponding uniformity property? In the case T(P) = P, the answer to the first questions is yes; the answer to the second is no. However, bootstrap confidence statements about T(P) can be made uniform over a restricted, though large, class of P. Similar results apply to other functional T(P).     

Approximate multivariate conditional inference using the adjusted profile likelihood

The author proposes saddlepoint approximation methods that are adapted to multivariate conditional inference in canonical exponential familles. Several approaches to approximating conditional discrete distributions involve dividing an approximation to the full joint mass function, summed over tail regions of interest, by an approximate marginal density. The author first approximates this conditional likelihood by the adjusted profile likelihood, and then applies a multivariate saddlepoint approximation. He also presents formulas to aid in performing simultaneously the profiling and maximizing steps.     

Assessing and comparing the accuracy of various bootstrap methods

We evaluate the performance of various bootstrap methods for constructing confidence intervals for mean and median of several common distributions. Using Monte Carlo simulation, we assessed performance by looking at coverage percentages and average confidence interval lengths. Poor performance is characterized by coverage deviating from 0.95 and large confidence interval lengths. Undercoverage is of greater concern than overcoverage. We also assess the performance of bootstrap methods in estimating the parameters of the Cox Proportional Hazard model and Accelerated Failure Time model.     

On the accuracy of the bootstrap approximation of the joint distribution of sample quantiles

It is proved that the accuracy of the bootstrap approximation of the joint distribution of sample quantiles lies between O(n^?1/4) and O(n^?1/4 a_n), where (log(n))1/2=O(a_n). As an application, we investigated confidence intervals based on the bootstrap.     

Simple and accurate inference for the mean of the gamma model

The two-parameter gamma model is widely used in reliability, environmental, medical and other areas of statistics. It has a two-dimensional sufficient statistic, and a two-dimensional parameter which can be taken to describe shape and mean. This makes it closely comparable to the normal model, but it differs substantially in that the exact distribution for the minimal sufficient statistic is not available. Some recently developed asymptotic theory is used to derive an approximation for observed levels of significance and confidence intervals for the mean parameter of the model. The approximation is as easy to apply as first-order methods, and substantially more accurate.     

On cornish-fisher expansions as a ratio of determinants without cumulants

A new method of approximating one quantile of a distribution function in terms of the corresponding quantile of another distribution function is introduced. The method utilizes the Cornish-Fisher expansion so as to eliminate the requirement for knowing the cumulants while at the same time retaining the desired simplicity as well as the property of not affecting the order of the error of the approximation.     

Implementation of saddlepoint approximations in resampling problems

In many situations saddlepoint approximations can replace the Monte Carlo simulation typically used to find the bootstrap distribution of a statistic. We explain how bootstrap and permutation distributions can be expressed as conditional distributions and how methods for linear programming and for fitting generalized linear models can be used to find saddlepoint approximations to these distributions. The ideas are illustrated using an example from insurance.     

Inference in the presence of redundant moment conditions and the impact of government health expenditure on health outcomes in England

In his 1999 article with Breusch, Qian, and Wyhowski in the Journal of Econometrics, Peter Schmidt introduced the concept of “redundant” moment conditions. Such conditions arise when estimation is based on moment conditions that are valid and can be divided into two subsets: one that identifies the parameters and another that provides no further information. Their framework highlights an important concept in the moment-based estimation literature, namely, that not all valid moment conditions need be informative about the parameters of interest. In this article, we demonstrate the empirical relevance of the concept in the context of the impact of government health expenditure on health outcomes in England. Using a simulation study calibrated to this data, we perform a comparative study of the finite performance of inference procedures based on the Generalized Method of Moment (GMM) and info-metric (IM) estimators. The results indicate that the properties of GMM procedures deteriorate as the number of redundant moment conditions increases; in contrast, the IM methods provide reliable point estimators, but the performance of associated inference techniques based on first order asymptotic theory, such as confidence intervals and overidentifying restriction tests, deteriorates as the number of redundant moment conditions increases. However, for IM methods, it is shown that bootstrap procedures can provide reliable inferences; we illustrate such methods when analysing the impact of government health expenditure on health outcomes in England.     

The use of the weighted likelihood in the natural exponential families with quadratic variance

The authors propose a weighted likelihood concept for the estimation of parameters in natural exponential families with quadratic variance. They apply the results to both simulated and real data.     

On the relative behavior of estimators of reliability/survivability

Based on a decomposition of mean absolute error, a twofold technique is introduced whereby a pairwise comparison of point estimators of reliability/survivability can be made. Given two such estimators, the method examines (a) the “odds” in favor of one of the estimators being closer to the true value than is the other and (b) each estimator’s average closeness to the true value not only when it is closer than is the other but also when it is not. Joint consideration of these concepts is shown to form a basis for determining which of the two estimators is preferred in a given situation. An application of the theory is made by comparing the maximum likelihood and minimum variance unbiased estimators of reliability/survivability in the exponential failure model.     

On a data-dependent choice of the tuning parameter appearing in certain goodness-of-fit tests

We propose a data-dependent method for choosing the tuning parameter appearing in many recently developed goodness-of-fit test statistics. The new method, based on the bootstrap, is applicable to a class of distributions for which the null distribution of the test statistic is independent of unknown parameters. No data-dependent choice for this parameter exists in the literature; typically, a fixed value for the parameter is chosen which can perform well for some alternatives, but poorly for others. The performance of the new method is investigated by means of a Monte Carlo study, employing three tests for exponentiality. It is found that the Monte Carlo power of these tests, using the data-dependent choice, compares favourably to the maximum achievable power for the tests calculated over a grid of values of the tuning parameter.