Proportional hazards models based on biased samples and estimated selection probabilities

In non‐randomized biomedical studies using the proportional hazards model, the data often constitute an unrepresentative sample of the underlying target population, which results in biased regression coefficients. The bias can be avoided by weighting included subjects by the inverse of their respective selection probabilities, as proposed by Horvitz & Thompson (1952) and extended to the proportional hazards setting for use in surveys by Binder (1992) and Lin (2000). In practice, the weights are often estimated and must be treated as such in order for the resulting inference to be accurate. The authors propose a two‐stage weighted proportional hazards model in which, at the first stage, weights are estimated through a logistic regression model fitted to a representative sample from the target population. At the second stage, a weighted Cox model is fitted to the biased sample. The authors propose estimators for the regression parameter and cumulative baseline hazard. They derive the asymptotic properties of the parameter estimators, accounting for the difference in the variance introduced by the randomness of the weights. They evaluate the accuracy of the asymptotic approximations in finite samples through simulation. They illustrate their approach in an analysis of renal transplant patients using data obtained from the Scientific Registry of Transplant Recipients     

Existence of approximate social welfare

This paper presents weak requirements on incomplete social preferences, and illustrates how such preferences can be derived from a choice rule. While the requirements do not guarantee the existence of a social welfare function, they do suffice for social preferences to be mapped faithfully to neighborhoods on the real line. That is, whenever consistent social choices are possible between given alternatives, the associated welfare intervals do not overlap.     

Reducing Palilalia by Presenting Tact Corrections to Young Children with Autism

Palilalia, the delayed repetition of words or phrases, occurs frequently among individuals with autism and developmental disabilities. The current study used a combined multiple baseline and reversal design to investigate the effectiveness of presenting tacts as corrections for palilalia. During baseline, five preschoolers with autism emitted high rates of palilalia and low rates of mands and tacts during play and instructional activities. During treatment, when experimenters presented opportunities to echoically tact actions and objects following the emission of palilalia, its frequency decreased to low and stable levels and mands and tacts increased. Functional relationships between the tact corrections and emissions of palilalia, mands, and tacts were established during reversals to baseline and treatment conditions. Similar trends in responding were found for frequency of palilalia, mands, and tacts in non-treatment settings.     

Robust prediction and extrapolation designs for misspecified generalized linear regression models

We study minimax robust designs for response prediction and extrapolation in biased linear regression models. We extend previous work of others by considering a nonlinear fitted regression response, by taking a rather general extrapolation space and, most significantly, by dropping all restrictions on the structure of the regressors. Several examples are discussed.     

A multi-resolution census algorithm for calculating vortex statistics in turbulent flows

Summary.  The fundamental equations that model turbulent flow do not provide much insight into the size and shape of observed turbulent structures. We investigate the efficient and accurate representation of structures in two-dimensional turbulence by applying statistical models directly to the simulated vorticity field. Rather than extract the coherent portion of the image from the background variation, as in the classical signal-plus-noise model, we present a model for individual vortices using the non-decimated discrete wavelet transform. A template image, which is supplied by the user, provides the features to be extracted from the vorticity field. By transforming the vortex template into the wavelet domain, specific characteristics that are present in the template, such as size and symmetry, are broken down into components that are associated with spatial frequencies. Multivariate multiple linear regression is used to fit the vortex template to the vorticity field in the wavelet domain. Since all levels of the template decomposition may be used to model each level in the field decomposition, the resulting model need not be identical to the template. Application to a vortex census algorithm that records quantities of interest (such as size, peak amplitude and circulation) as the vorticity field evolves is given. The multiresolution census algorithm extracts coherent structures of all shapes and sizes in simulated vorticity fields and can reproduce known physical scaling laws when processing a set of vorticity fields that evolve over time.     

Treating distressed couples with coexisting mental and physical disorders: directions for clinical training and practice

Relationship difficulties frequently coexist with significant emotional, behavioral, and physical health problems in one or both partners. Moreover, couple therapists report individual emotional and behavioral disorders as among the most difficult problems that they encounter in clinical practice. In this article, we summarize empirical evidence regarding the association between relationship distress and the onset, course, and treatment of mental and physical health problems. Based on these findings, we propose five critical implications for clinical training and the practice of couple therapy.     

Common factors and our sacred models

In this article we argue that much of what makes one treatment effective is common to other forms of effective treatment--both in psychotherapy generally and in marital and family therapy (MFT) specifically. Yet MFT has largely ignored the research on common factors. In this article we present a moderate view of common factors that, while repudiating the extreme position that there is no difference among treatment models, stresses that there are common factors and mechanisms of change that undergird most forms of successful treatment. These common mechanisms of change should be given more attention in our field, which has tended to emphasize the uniqueness of our sacred models. We delineate some of the major common factors, review the empirical evidence for them, and discuss implications of adapting a common factors informed approach to family therapy.     

Gender, widowhood, and long-term care in the older Mexican American population

The objective of this study is to examine the influences of gender on long-term care service use among older Mexican American widows and widowers. Our analysis is based on a sample of 773 widows and 183 widowers from the Longitudinal Study of Elderly Mexican American Health (H-EPESE). In this sample widows resemble widowers in terms of demographic and health characteristics. However, widows report more financial strain than widowers and a greater welfare dependency (SSI) and Medicaid use. Among those who suffered diminished health, widows were more likely than widowers to use community-based long-term care services whereas widowers were more likely to enter a nursing home. Widows also had more instrumental and socioemotional support than widowers. Serious cognitive and functional impairment, though, places widows and widowers at the same risk of institutionalization. We end with a discussion of the policy implications of these findings.     

Outcome Expectancies for Gambling: Empirical Modeling of a Memory Network in Federal Prison Inmates

Outcome expectancies for gambling were explored in a group of 316 male medium security federal inmates with the multidimensional scaling statistical technique. Sixty-six possible outcomes for gambling were reduced to 16 after a factor analysis determined that 75% of the total variance was accounted for by the first 16 factors (eigenvalues > 0.75). The highest loading item for each factor was utilized in all subsequent analyses. Multidimensional scaling of these 16 items revealed the presence of two primary dimensions (positive-negative and arousing-sedating), corroborating previous research on alcohol outcome expectancies (Rather et al., 1992) and general affective response (Russell, 1980). Significant group (non-gambling, non-problem gambling, possible problem gambling, probable pathological gambling) differences were observed on 12 of the 16 items and on 3 of 4 composite scales (positive expectancies, negative expectancies, arousing expectancies, sedating expectancies).