Proposed best practice for statisticians in the reporting and publication of pharmaceutical industry‐sponsored clinical trials

In this paper we set out what we consider to be a set of best practices for statisticians in the reporting of pharmaceutical industry‐sponsored clinical trials. We make eight recommendations covering: author responsibilities and recognition; publication timing; conflicts of interest; freedom to act; full author access to data; trial registration and independent review. These recommendations are made in the context of the prominent role played by statisticians in the design, conduct, analysis and reporting of pharmaceutical sponsored trials and the perception of the reporting of these trials in the wider community. Copyright © 2010 John Wiley & Sons, Ltd.     

Response transformations in repeated measures and growth curve models

This article describes estimation and inference procedures for the parameters of the Box-Cox and foided-power transformations in repeated measures and growth curve models. Procedures for computing maximum likelihood estimates of the transformation and covariance parameters under several covanance structures (omnibus sphericity, local sphericity, and unstructured) are described. Lack of fit statistics and hypothesis tests for comparing these structures also are described. The procedures are illustrated on three data sets. Software for performing the analyses in the SAS System is described and is available from the authors.     

Mexican Immigrants,Labour Market Assimilation and the Current Population: The Sensitivity of Results Across Seemingly Equivalent Surveys

In this paper we compare estimates of immigrants’ labour supply assimilation profiles using the Current Population Survey Annual Demographic Files (March ADS) and the Current Population Survey Outgoing Rotation Groups (ORGs). We use a measure that is seemingly consistent across both surveys: usual weekly hours of work in the main job. This measure is not only important as it measures the intensive margin of labour supply, but it is also used to estimate hourly earnings in both surveys. Our results indicate that the two surveys produce dramatically different estimates of the change in average hours of work as immigrants’ years in the United States increase: estimates from the March ADS predict much steeper hour's assimilation profiles than do estimates obtained from the ORGs. These differences persist if we include controls for demographic characteristics, and only disappear once we control for occupation heterogeneity. We argue that these differences stem from two separate factors that differentiate the data. First, the ADS and ORG frame the usual “hours worked” question differently. Also, differences in the timing of the surveys may produce seasonality effects that differentially affect the composition of recent and earlier migrants, thereby changing assimilation profiles.     

Men's sexual coercion in intimate relationships: development and initial validation of the Sexual Coercion in Intimate Relationships scale

We report the development and initial validation of the Sexual Coercion in Intimate Relationships Scale (SCIRS), a measure designed to assess the prevalence and severity of sexual coercion in committed intimate relationships. We review existing measures of sexual coercion and discuss their limitations, describe the identification of the SCIRS items, perform a principal components analysis and describe the resulting three components, and present evidence for the convergent and discriminative validity of the SCIRS. Because sexual coercion in intimate relationships often takes the form of subtle tactics, the SCIRS items assess communicative tactics such as hinting and subtle manipulations in addition to tactics such as the use of physical force. The SCIRS provides researchers and clinicians with a valid, reliable, and comprehensive measure with which to study the dynamics of sexual coercion in intimate relationships.     

Couple age discrepancy and risk of intimate partner homicide

Although national level studies in the United States and Canada find that extreme partner age discrepancy is a risk factor for intimate partner homicide in opposite-sex couples, these studies carry two caveats: They are limited to cohabiting marital or common-law couples and they are not detailed enough to explore alternative explanations for the age discrepancy-homicide risk association. Using the Chicago Homicide Dataset, which includes all homicides that occurred in Chicago from 1965 to 1996, we analyze the 2,577 homicides in which the victim was killed by a current or former legal spouse, common-law spouse, or heterosexual boyfriend or girlfriend, and in which the woman was at least 18 years of age. Within each of 14 categories of couple age discrepancy, we estimate the population of intimate heterosexual couples and calculate the population-based risk of homicide. The results replicate national level findings showing that the risk of intimate partner homicide is considerably elevated for couples with a large discrepancy between their ages-where the man is at least 16 years older than the woman or the woman is at least 10 years older than the man. This risk pattern occurs regardless of whether the man or the woman was the homicide offender. We then investigate whether the link between partner age discrepancy and homicide risk is explained by the offender's arrest record. Results show that the higher risk of intimate partner homicide for age discrepant couples is robust, and does not depend on the previous arrest record of the offender. Discussion addresses other possible explanations for the increased risk of partner homicide for age discrepant couples, and the practical implications of these findings.     

Locating Zombies in the Sociology of Popular Culture

In this essay, focusing primarily on the cinema of the walking corpse, I provide an overview of zombie studies and suggest potential avenues for sociological inquiry into zombie phenomena. I argue that zombie films, comic books, novels, video games, and the like can be seen as significant cultural objects that reflect and reveal the cultural and material circumstances of their creation. Despite emanating from complex culture‐producing institutions and (arguably) capturing extant social anxieties, sociology has remained quiet on zombie phenomena. Issues of significance, history, and definition are discussed. I then locate three avenues of inquiry ideally suited to the sociological toolkit: symptomatic analysis of content, production, and audience response and interaction. I conclude by calling for a multipronged sociological analysis into “zombie culture.”     

A Psychometric Analysis of Field Evaluation Instruments Designed to Measure Students’ Generalist-Level Social Work Competencies

ABSTRACT

There is a dearth of research on the evaluation of the psychometric performance of instruments designed to measure students’ generalist-level social work competencies. There is also uncertainty on the performance of various response option formats used to measure students’ competencies in assessment instruments. Using a sample of 198 master of social work students and 198 field instructors, the current study employs confirmatory factor analysis to assess the psychometric performance of a field evaluation instrument designed to measure students’ generalist-level social work competencies. The results provide evidence to support the validity and reliability of the instrument. The results also highlight nuances associated with a 5-point response option format for items used to measure competency indicators. Implications, limitations, and directions for future research are discussed.     

Use of Multicriteria Decision Analysis to Support Weight of Evidence Evaluation

Weight of evidence (WOE) methods are key components of ecological and human health risk assessments. Most WOE applications rely on the qualitative integration of diverse lines of evidence (LOE) representing impact on ecological receptors and humans. Recent calls for transparency in assessments and justifiability of management decisions are pushing the community to consider quantitative methods for integrated risk assessment and management. This article compares and contrasts the type of information required for application of individual WOE techniques and the outcomes that they provide in ecological risk assessment and proposes a multicriteria decision analysis (MCDA) framework for integrating individual LOE in support of management decisions. The use of quantitative WOE techniques is illustrated for a hypothetical but realistic case study of selecting remedial alternatives at a contaminated aquatic site. Use of formal MCDA does not necessarily eliminate biases and judgment calls necessary for selecting remedial alternatives, but allows for transparent evaluation and fusion of individual LOE. It also provides justifiable methods for selecting remedial alternatives consistent with stakeholder and decision‐maker values.     

Marginal and hazard ratio specific random data generation: Applications to semi-parametric bootstrapping

Cox's partial likelihood for censored time-to-event data can be interpreted as a permutation probability, whereby covariate values are permuted to the observed times-to-event and censoring times. This interpretation facilitates a simple method for jointly generating times-to-event and covariate tuples with considerable flexibility, including time dependence of the hazard ratio and specification of both the marginal time-to-event and covariate distributions. This interpretation also facilitates a method for semi-parametric bootstrapping of hazard ratio estimators.