Cox regression with missing covariate data using a modified partial likelihood method

Missing covariate values is a common problem in survival analysis. In this paper we propose a novel method for the Cox regression model that is close to maximum likelihood but avoids the use of the EM-algorithm. It exploits that the observed hazard function is multiplicative in the baseline hazard function with the idea being to profile out this function before carrying out the estimation of the parameter of interest. In this step one uses a Breslow type estimator to estimate the cumulative baseline hazard function. We focus on the situation where the observed covariates are categorical which allows us to calculate estimators without having to assume anything about the distribution of the covariates. We show that the proposed estimator is consistent and asymptotically normal, and derive a consistent estimator of the variance–covariance matrix that does not involve any choice of a perturbation parameter. Moderate sample size performance of the estimators is investigated via simulation and by application to a real data example.     

Calibration estimation in dual-frame surveys

Survey statisticians make use of auxiliary information to improve estimates. One important example is calibration estimation, which constructs new weights that match benchmark constraints on auxiliary variables while remaining “close” to the design weights. Multiple-frame surveys are increasingly used by statistical agencies and private organizations to reduce sampling costs and/or avoid frame undercoverage errors. Several ways of combining estimates derived from such frames have been proposed elsewhere; in this paper, we extend the calibration paradigm, previously used for single-frame surveys, to calculate the total value of a variable of interest in a dual-frame survey. Calibration is a general tool that allows to include auxiliary information from two frames. It also incorporates, as a special case, certain dual-frame estimators that have been proposed previously. The theoretical properties of our class of estimators are derived and discussed, and simulation studies conducted to compare the efficiency of the procedure, using different sets of auxiliary variables. Finally, the proposed methodology is applied to real data obtained from the Barometer of Culture of Andalusia survey.     

Finie mixures of canonical fundamenal skew $$$$-disribuions

This paper introduces a finite mixture of canonical fundamental skew \(t\) (CFUST) distributions for a model-based approach to clustering where the clusters are asymmetric and possibly long-tailed (in: Lee and McLachlan, arXiv:1401.8182 [statME], 2014b). The family of CFUST distributions includes the restricted multivariate skew \(t\) and unrestricted multivariate skew \(t\) distributions as special cases. In recent years, a few versions of the multivariate skew \(t\) (MST) mixture model have been put forward, together with various EM-type algorithms for parameter estimation. These formulations adopted either a restricted or unrestricted characterization for their MST densities. In this paper, we examine a natural generalization of these developments, employing the CFUST distribution as the parametric family for the component distributions, and point out that the restricted and unrestricted characterizations can be unified under this general formulation. We show that an exact implementation of the EM algorithm can be achieved for the CFUST distribution and mixtures of this distribution, and present some new analytical results for a conditional expectation involved in the E-step.     

Variable selection for survival data with a class of adaptive elastic net techniques

The accelerated failure time (AFT) models have proved useful in many contexts, though heavy censoring (as for example in cancer survival) and high dimensionality (as for example in microarray data) cause difficulties for model fitting and model selection. We propose new approaches to variable selection for censored data, based on AFT models optimized using regularized weighted least squares. The regularized technique uses a mixture of \(\ell _1\) and \(\ell _2\) norm penalties under two proposed elastic net type approaches. One is the adaptive elastic net and the other is weighted elastic net. The approaches extend the original approaches proposed by Ghosh (Adaptive elastic net: an improvement of elastic net to achieve oracle properties, Technical Reports 2007) and Hong and Zhang (Math Model Nat Phenom 5(3):115–133 2010), respectively. We also extend the two proposed approaches by adding censoring observations as constraints into their model optimization frameworks. The approaches are evaluated on microarray and by simulation. We compare the performance of these approaches with six other variable selection techniques-three are generally used for censored data and the other three are correlation-based greedy methods used for high-dimensional data.     

Eminent Social Scientists’ Views on Immigration Issues: 1985 and 2006

Following up on a research note by Stephen Moore published in 1985, this article compares the results of the survey conducted in 1985 against one carried out in 2006 on the views of eminent social scientists (economists, political scientists, psychologists, sociologists, historians and anthropologists) on immigration issues. Specifically, the surveys report on whether the respondents are likely to favor an increase in the number of immigrants who are allowed into the United States and other issues such as whether immigrants improve the economy.

---

Passing Encounters East and West: Comparing Japanese and American Pedestrian Interactions

This study examined the microinteractions of pedestrians in Japan and in the United States as they walked past a confederate. Specifically, the effects of culture, condition (avoid, look-only, and look plus smile) and sex of confederate on glances, smiles, nods, and greetings by passing pedestrians were examined in a field study on over 1000 participants. The hypotheses of (1) lower responsiveness in Japanese pedestrians than in American pedestrians and (2) increased responsiveness as a function of condition were supported in a series of log-linear analyses of pedestrian glances, smiles, nods, and greetings. Both of these main effects were, however, qualified by Culture X Condition interactions on smiles, nods, and greetings, with the large condition effects present in the American pedestrians, but absent in the Japanese pedestrians. The results are discussed in terms of the functions of glances, smiles, nods, and greetings in these brief encounters and how differing cultural norms affect Japanese and American responsiveness. Finally, the limitations of this study and the broader utility of this research paradigm are discussed.

Injunctive Norms and Problem Gambling among College Students

Two studies examined the relationships among injunctive norms and college student gambling. In study 1 we evaluated the accuracy of perceptions of other students’ approval of gambling and the relationship between perceived approval and gambling behavior. In study 2 we evaluated gambling behavior as a function of perceptions of approval of other students, friends, and family. In study 1, which included 2524 college students, perceptions of other students’ approval of gambling were found to be overestimated and were negatively associated with gambling behavior. The results of study 2, which included 565 college students, replicated the findings of study 1 and revealed positive associations between gambling behavior and perceived approval of friends and family. Results highlight the complexity of injunctive norms and the importance of considering the reference group (e.g., peers, friends, family members) in their evaluation. Results also encourage caution in considering the incorporation of injunctive norms in prevention and intervention approaches.     

Gambling and Alcohol Use: Trends in the State of New Mexico from 1996–1998

Alcohol consumption and its relationship to gambling was examined in a statewide New Mexico survey in 1996 and 1998. Data regarding both drinking habits and gambling behavior were obtained from a stratified random sample of the adult population (N = 2674) across the entire State of New Mexico via phone survey. These surveys were carried out shortly after a period when New Mexico experienced an initial surge in the legalized gaming industry. Fifty-seven percent of survey respondents reported drinking in the past month, while 43.1% reported no drinking in the past 30 days. Results show that while the number of days in the past 30 that a person drinks is significantly correlated with some types of gambling behavior (e.g., in 1998, drinking more days was associated with more card gambling and sports gambling), it is the amount of alcohol consumed per occasion that is associated with more gambling behavior. For example, in both 1996 and 1998, drinking more per occasion was associated with more sports betting, dice gambling, number/lottery gambling, gambling using machines, and paper game gambling (e.g., pull tabs, punchboard). These results suggest that problem gambling behavior is not affected as much by the number of occasions on which one consumes alcohol, as by the amount of alcohol consumed per occasion.     

Gambling Behavior of Student-Athletes and a Student Cohort: What are the Odds?

This study investigated the prevalence of gambling, gambling related NCAA violations, and disordered gambling in student-athletes (n = 736) with a comparison cohort of students (n = 1,071) at four universities. Student-athletes reported similar rates of gambling frequency, use of a bookmaker, and disordered gambling as students. After accounting for demographic differences, student-athletes were less likely to engage in sports wagering than students. Several risk factors for disordered gambling were identified, including being male and reporting at least one parent with a history of gambling problems. These findings suggest that problems associated with gambling are a university-wide issue with student-athletes meriting additional attention because of implications for the integrity of intercollegiate sports. Improved prevention and intervention efforts for collegiate gambling are recommended.     

Self-perception of Gambling Problems among Adolescents Identified as At-risk or Problem Gamblers

The relative influence of perceived familial addictive behaviors and personal gambling behaviors on adolescents' self-perceptions of gambling problems was examined. Students from five high schools in Connecticut (N = 3,886) were surveyed. Of those between the ages of 14 and 17 who scored two or more on the South Oaks Gambling Screen--Revised for Adolescents (n = 532; 72% male; 43% Caucasian), 14.3% reported having a current or past problem with gambling. Wagering larger amounts in a single day, gambling on a daily basis, and perceived presence of a family member with a gambling problem were associated with increased odds of self-perception of a gambling problem. Thus, adolescents who may be less likely to be identified for prevention efforts (due to lack of engagement in high stakes gambling or the real/perceived absence of a problematic gambler in the home) appear less likely to perceive a gambling problem. To advance prevention and treatment strategies, the apparent discrepancy between adolescents' self-perceptions and objective reports of problem gambling behaviors warrants further investigation.