Problems with the measures of gastrointestinal and cardiovascular symptom frequency in The Standard Shiftwork Index

The measures of gastrointestinal and cardiovascular health in The Standard Shiftwork Index (SSI; Barton, Costa, Smith, Spelten, Totterdell, and Folkard, 1995) purport to measure the frequency of 19 related symptoms using a 4-point rating scale, with response categories expressed in terms of a relative frequency: 'almost never', 'quite seldom', 'quite often', and 'almost always'. The present pilot study examined the relationship between frequency ratings of 11 gastrointestinal and 8 cardiovascular symptoms made on the SSI 4-point scale, and ratings of the same symptoms made on an alternative 9-point rating scale, which had response categories that were expressed in terms of an absolute value of symptom frequency (e.g. 'once a week'). Results obtained from a sample of 56 non-shiftworkers showed that the correlations between the two response formats were not high for any of the 11 gastrointestinal or 8 cardiovascular symptoms. Three of the relationships were non-significant. In addition, substantial variability in the interpretation of the response categories was detected between symptom items pertaining to both the gastrointestinal and cardiovascular subscales. The implications of these findings are discussed in relation to the current system of scoring individual symptom items, and the current aims within shiftwork research.     

An estimating equation for parametric shared frailty models with marginal additive hazards

Summary.  Multivariate failure time data arise when data consist of clusters in which the failure times may be dependent. A popular approach to such data is the marginal proportional hazards model with estimation under the working independence assumption. In some contexts, however, it may be more reasonable to use the marginal additive hazards model. We derive asymptotic properties of the Lin and Ying estimators for the marginal additive hazards model for multivariate failure time data. Furthermore we suggest estimating equations for the regression parameters and association parameters in parametric shared frailty models with marginal additive hazards by using the Lin and Ying estimators. We give the large sample properties of the estimators arising from these estimating equations and investigate their small sample properties by Monte Carlo simulation. A real example is provided for illustration.     

Employee strain and job satisfaction related to an implementation of quality in a public service organization: a longitudinal study

To date, few studies have focused on employee reactions to the quality-related aspects of the introduction of New Public Management (NPM). The aim of this study was to investigate the effects on employee strain and satisfaction of the implementation NPM in a public service organization. The study was designed as an empirical examination of the 'context-dependent approach' ( Edwards, Collinson, & Rees, 1998 ). This approach suggests that effects of an organizational change depend on the context of the implementation, i.e. organizational aspects (implementation strategies), job content and job context dimensions. They employed a single-case longitudinal design. The case was a large municipal service unit responsible for the public housing system of a city in Austria. The sample consisted of 217 employees. Measurements were taken before the organizational change, and at two different times during the change process. While the implementation of NPM can be considered to have been an organizational success (increase in customer satisfaction), it was accompanied by increases in job strain, and, at the same time, mixed results in job satisfaction. Less qualified employees mainly responsible for the customer interface experienced the organizational change most negatively. Using structural equation modelling, the context-dependent approach was empirically confirmed. Job control, role clarity and information were found to be the most important job and organizational resources.     

A Likelihood Based Estimating Equation for the Clayton–Oakes Model with Marginal Proportional Hazards

Multivariate failure time data arise when data consist of clusters in which the failure times may be dependent. A popular approach to such data is the marginal proportional hazards model with estimation under the working independence assumption. In this paper, we consider the Clayton–Oakes model with marginal proportional hazards and use the full model structure to improve on efficiency compared with the independence analysis. We derive a likelihood based estimating equation for the regression parameters as well as for the correlation parameter of the model. We give the large sample properties of the estimators arising from this estimating equation. Finally, we investigate the small sample properties of the estimators through Monte Carlo simulations.     

„Public Sociologies“

Ohne Zusammenfassung     

Une condition nécessaire d'admissibilité et ses conséquences sur les estimateurs à rétrécisseur de la moyenne d'un vecteur normal

Considering exponential families of distributions, we estimate parameters which are not the natural parameters. We prove that the admissible estimators of these parameters are limits of Bayes estimators and can be expressed through a given functional form. An important particular case of this model pertains to the estimation of the mean of a multidimensional normal distribution when the variance is known up to a multiplicative factor. We deduce from the main result a necessry condition for the admissibility of matricial shrinkage estimators.     

Kriminalitätsfurcht im urbanen raum

Ziel ist die Erkl?rung von Kriminalit?tsfurcht auf verschiedenen Dimensionen (affektiv, kognitiv, konativ). Es werden drei theoretische Modelle zur Erkl?rung von Kriminalit?tsfurcht empirisch überprüft (Viktimisierungsmodell, Disorder-Modell, Modell sozialer Integration). Auf der Grundlage einer postalischen Bev?lkerungsumfrage (N = 3.612) in 49 Hamburger Stadtteilen werden Mehrebenenmodelle getestet. Dabei zeigt sich, dass die Kriminalit?tsfurcht zwischen den Stadtteilen signifikant variiert. In der Mehrebenenanalyse best?tigen sich Effekte von Individualvariablen (perzipierte physical und social incivilities, pers?nliche und indirekte Viktimisierungen, Vertrauen zu Nachbarn, perzipierte Polizeistreifen, Alter, Geschlecht) und Kontextvariablen (problematische Sozialstruktur, Bev?lkerungsdichte) auf die verschiedenen Dimensionen der Kriminalit?tsfurcht.