Measuring the complexity of generalized linear hierarchical models

Measuring a statistical model's complexity is important for model criticism and comparison. However, it is unclear how to do this for hierarchical models due to uncertainty about how to count the random effects. The authors develop a complexity measure for generalized linear hierarchical models based on linear model theory. They demonstrate the new measure for binomial and Poisson observables modeled using various hierarchical structures, including a longitudinal model and an areal‐data model having both spatial clustering and pure heterogeneity random effects. They compare their new measure to a Bayesian index of model complexity, the effective number p_D of parameters (Spiegelhalter, Best, Carlin & van der Linde 2002); the comparisons are made in the binomial and Poisson cases via simulation and two real data examples. The two measures are usually close, but differ markedly in some instances where p_D is arguably inappropriate. Finally, the authors show how the new measure can be used to approach the difficult task of specifying prior distributions for variance components, and in the process cast further doubt on the commonly‐used vague inverse gamma prior.     

Movers and Stayers in Care Management in Adult Services

Professional biographies of care managers in the mid-ninetiesand, seven years on, their profiles and career choice are thefocus of this article. The work draws on two linked studiesbased on adult services in two social services departments.The profile of those who stayed in post and those who movedon raises issues for the social work profession, for managersand for policy makers. Those who stayed constitute an experiencedworkforce of middle-aged, predominantly female practitionerswith mixed views about the changing nature of the work. Thosewho moved include a group of females who became managers bydefault. The work raises issues concerning work pressures, genderprofiles within an ageing workforce, and succession planning.The findings link to more extensive workforce studies and contributetowards debates concerning the occupational identity and thefuture of qualified practitioners within a human services landscapewhich is constantly being reshaped within a culture of new managerialism.     

Outcomes assessment of role-play scenarios for teaching responsible conduct of research

We describe the summative assessment of role-play scenarios that we previously developed to teach central topics in the responsible conduct of research (RCR) to graduate students in science and engineering. Interviews with role-play participants, with participants in a case discussion training session, and with untrained students suggested that role-playing might promote a deeper appreciation of RCR by shifting the focus away from wanting to simply "know the rules." We also present the results of a think-aloud case analysis study and describe the development of a behaviorally-anchored rating scale (BARS) to assess participants' case analysis performance.     

Outcomes Assessment of Role-Play Scenarios for Teaching Responsible Conduct of Research

We describe the summative assessment of role-play scenarios that we previously developed to teach central topics in the responsible conduct of research (RCR) to graduate students in science and engineering. Interviews with role-play participants, with participants in a case discussion training session, and with untrained students suggested that role-playing might promote a deeper appreciation of RCR by shifting the focus away from wanting to simply “know the rules.” We also present the results of a think-aloud case analysis study and describe the development of a behaviorally-anchored rating scale (BARS) to assess participants' case analysis performance.     

An iterative Monte Carlo method for nonconjugate Bayesian analysis

The Gibbs sampler has been proposed as a general method for Bayesian calculation in Gelfand and Smith (1990). However, the predominance of experience to date resides in applications assuming conjugacy where implementation is reasonably straightforward. This paper describes a tailored approximate rejection method approach for implementation of the Gibbs sampler when nonconjugate structure is present. Several challenging applications are presented for illustration.     

The impact of involuntary job loss on those disabled by society: a pilot study to encourage effective participation

Over half of the people who become disabled whilst employed are thought to be pressured into leaving their job. This study investigates the experience of involuntary retirement/redundancy due to disability. Three researchers who had all experienced such involuntary job loss were trained to conduct the research. Semi-structured interviews were designed and conducted by disabled researchers. The training of the researchers enabled them to successfully conduct a research project. The results of the project show the impact of involuntary job loss due to disability to be considerable. This study would suggest that government reforms to create new opportunities for disabled people of a working age are well overdue. Although there were some problems experienced with the amount of training required and the lack of objectivity, training disabled researchers to conduct research proved to be a valuable experience for both the researchers and the facilitators.     

A SEMI-PARAMETRIC MODEL FOR ASYMMETRIC INFORMATION IN HEALTH PLAN MARKETS

Two sources of asymmetric information in health markets are adverse selection where only individuals know their illness probability and their preventive effort to reduce illness probabilities. This effort is not observed by insurers, and thus individuals cannot be adequately compensated for their effort. I investigate whether asymmetric information plays a role in health care markets through a semi-parametric model that allows both adverse selection and preventive effort. I find that preventive effort plays a role in illness probabilities. If the United States implements universal health coverage, the currently uninsured population will exert less preventive effort, and their illness probabilities will increase. (JEL I11 , C14 )     

Exploring Nonnormative Coresidence in Urban China: Living With Wives’ Parents

We model histories between two cohorts of urban Chinese couples (N = 1,191) of a rarely studied living arrangement—coresidence with the wife's parents—using a dynamic life history analysis in contrast to previous cross‐sectional studies of coresidence. We examine patterns of entry into and exit from coresidence with the wife's parents, comparing the predictive power of modernization theory to the effect of demographic change and the resources and needs of each generation. Given China's well‐known patrilineal family system, we find a surprisingly high number of couples ever residing nonnormatively, and significant differences between cohorts in what determines the pattern of coresidence. Resources and needs that reflect conscious choices to coreside most strongly influence nonnormative coresidence. Its importance may increase as the children of the One‐Child Family Policy grow up and marry.