Predicting random effects with an expanded finite population mixed model

Prediction of random effects is an important problem with expanding applications. In the simplest context, the problem corresponds to prediction of the latent value (the mean) of a realized cluster selected via two-stage sampling. Recently, Stanek and Singer [Predicting random effects from finite population clustered samples with response error. J. Amer. Statist. Assoc. 99, 119–130] developed best linear unbiased predictors (BLUP) under a finite population mixed model that outperform BLUPs from mixed models and superpopulation models. Their setup, however, does not allow for unequally sized clusters. To overcome this drawback, we consider an expanded finite population mixed model based on a larger set of random variables that span a higher dimensional space than those typically applied to such problems. We show that BLUPs for linear combinations of the realized cluster means derived under such a model have considerably smaller mean squared error (MSE) than those obtained from mixed models, superpopulation models, and finite population mixed models. We motivate our general approach by an example developed for two-stage cluster sampling and show that it faithfully captures the stochastic aspects of sampling in the problem. We also consider simulation studies to illustrate the increased accuracy of the BLUP obtained under the expanded finite population mixed model.     

MODEL SELECTION CRITERIA FOR LOGLINEAR MODELS

Considerable work has been devoted to developing model selection criteria for normal theory regression models. Less attention has been paid to discrete data. We develop two loglinear model selection criteria for Poisson counts. These criteria are based on an estimated bias adjustment of the Akaike information criterion. We observe in a simulation study that the corrected statistics provide good model choices and relatively accurate estimates of the mean structure.     

Empirical Bayes vs. fully Bayes variable selection

For the problem of variable selection for the normal linear model, fixed penalty selection criteria such as AIC, _Cp${\mathrm{C}}_p$, BIC and RIC correspond to the posterior modes of a hierarchical Bayes model for various fixed hyperparameter settings. Adaptive selection criteria obtained by empirical Bayes estimation of the hyperparameters have been shown by George and Foster [2000. Calibration and Empirical Bayes variable selection. Biometrika 87(4), 731–747] to improve on these fixed selection criteria. In this paper, we study the potential of alternative fully Bayes methods, which instead margin out the hyperparameters with respect to prior distributions. Several structured prior formulations are considered for which fully Bayes selection and estimation methods are obtained. Analytical and simulation comparisons with empirical Bayes counterparts are studied.     

Palliative healthcare: cost reduction and quality enhancement using end-of-life survey methodology

American medical institutions throughout the 20th century prescribed high customer satisfaction, but when it came to death, largely ignored it. An accelerated accumulation of esoteric medical information and the application of this knowledge to affect new cures and longer lives instilled an unquestioning reverence for the medical community among the patient population. Diminishing marginal gains in life expectancy, escalating costs related to life sustaining technologies, and a psychographic shift in the dominant consumer base have challenged this traditional reverence. Armed with unprecedented access to medical information, a more knowledgeable and assertive patient population has emerged in the 21st century to institute its own standards of what constitutes quality health care. In terms of end of life care, this has meant recognition that the emotional needs of the dying have been largely underserved by the current American medical model. Patients and their families are no longer willing to accept the traditional medical perspective of death as failure and have numerous international palliative care models that serve as benchmarks of success when it comes to quality of dying. When cure is a possibility, Americans will pursue it at all costs, but when it is not a possibility, they want honest communication and the opportunity to say good-bye to their loved ones. In the context of these emergent needs, life review is offered as a solution. The value proposition targets not only dying patients and their families, but also society as a whole.     

Non-native plants in the understory of riparian forests across a land use gradient in the Southeast

As urbanization expands into rural areas, an increase in the number of non-native plant species at the urban-rural interface is expected due in large part to the increased availability of propagules from ornamental plantings. A study investigating the distribution of non-native plants in the understories of riparian forests across an urban-to-rural gradient north of Columbus, GA was initiated in 2003. A significantly greater number of non-native plant species occurred at the urban sites and at one site at the urban-rural interface, where 20 to 33% of the species encountered were non-native. In contrast, at the more rural sites non-native species comprised 4–14% of the total number of species. However, the importance values of non-native species as a whole did not change significantly across the land use gradient due to the high frequency and abundance of three non-native species (Ligustrum sinense, Lonicera japonica, and Microstegium vimineum) in the majority of the watersheds. Reductions in species richness and overstory reproduction associated with these non-natives could impact long-term forest structure and ecosystem function.     

Development of a lateral mobility task to identify individuals at risk for mobility disability and functional decline

Lateral mobility is integral to many activities of daily living involving transfer from one position to another. The objective of this study was to develop and evaluate the validity and test-retest reliability of a lateral-mobility (LATMOB) task for older adults. Measurements of lateral mobility, balance, and strength and self-reported and performance-based physical functioning were obtained in 63 women and 77 men > or = 50 years of age. The LATMOB task was significantly correlated with age, knee-extensor strength, grip strength, functional reach, and one-leg-stance time. Test-retest reliability of the task was excellent. The LATMOB task was highly correlated with the car task. Balance was significantly correlated with time to get into and out of a car and performance on the LATMOB task. The LATMOB task was significantly correlated with the Short Physical Performance Battery score. The LATMOB task is valid and reliable, but additional work is needed to assess its sensitivity to change and predictive validity.     

Social security reform in the twenty-first century: the United States

The paper reviews the history of the Social Security system in the United States in the twentieth century and discusses options for the twenty-first. Because of the steady aging of the U.S. population and the impending retirement of the large baby boom cohort, the Social Security program now is in long-term actuarial deficit. The standard twentieth century approach to this actuarial deficit would be to raise payroll taxes enough to pay for anticipated future benefit increases, but for several reasons that approach may not be so popular this time around. The author's preferred approach is a gradual trimming of long-term benefit growth, plus "add on" individual accounts to provide new saving, for the economy and for the retirement system. The paper also criticizes proposals for Social Security reform made by President Clinton and a committee appointed by President Bush, generally because these proposals do not provide enough new saving.