An omnibus lack of fit test in logistic regression with sparse data

The usefulness of logistic regression depends to a great extent on the correct specification of the relation between a binary response and characteristics of the unit on which the response is recoded. Currently used methods for testing for misspecification (lack of fit) of a proposed logistic regression model do not perform well when a data set contains almost as many distinct covariate vectors as experimental units, a condition referred to as sparsity. A new algorithm for grouping sparse data to create pseudo replicates and using them to test for lack of fit is developed. A simulation study illustrates settings in which the new test is superior to existing ones. Analysis of a dataset consisting of the ages of menarche of Warsaw girls is also used to compare the new and existing lack of fit tests.     

Lung Cancer Risk from Radon in Marcellus Shale Gas in Northeast U.S. Homes

The amount of radon in natural gas varies with its source. Little has been published about the radon from shale gas to date, making estimates of its impact on radon‐induced lung cancer speculative. We measured radon in natural gas pipelines carrying gas from the Marcellus Shale in Pennsylvania and West Virginia. Radon concentrations ranged from 1,520 to 2,750 Bq/m³ (41–74 pCi/L), and the throughput‐weighted average was 1,983 Bq/m³ (54 pCi/L). Potential radon exposure due to the use of Marcellus Shale gas for cooking and space heating using vent‐free heaters or gas ranges in northeastern U.S. homes and apartments was assessed. Though the measured radon concentrations are higher than what has been previously reported, it is unlikely that exposure from natural gas cooking would exceed 1.2 Bq/m³ (<1% of the U.S. Environmental Protection Agency's action level). Using worst‐case assumptions, we estimate the excess lifetime (70 years) lung cancer risk associated with cooking to be 1.8×10^?4 (interval spanning 95% of simulation results: 8.5×10^?5, 3.4×10^?4). The risk profile for supplemental heating with unvented gas appliances is similar. Individuals using unvented gas appliances to provide primary heating may face lifetime risks as high as 3.9×10^?3. Under current housing stock and gas consumption assumptions, expected levels of residential radon exposure due to unvented combustion of Marcellus Shale natural gas in the Northeast United States do not result in a detectable change in the lung cancer death rates.     

American High School Students’ Perceptions of Determinants of Life Satisfaction

Life satisfaction can be assessed either globally or with regard to satisfaction with specific domains of life. The latter multidimensional approach presumes science has delineated with confidence the specific domains most relevant to evaluating whether the criteria for a good life have been met. This paper shares results of a qualitative study of the perceived determinants of life satisfaction among 30 high school students who were diverse in terms of mental health; 6–10 participants were classified as complete mental health, vulnerable, symptomatic but content, or troubled at two time points separated by a year. Thematic analyses of transcribed individual interviews suggested eight themes that capture the domains of life adolescents perceive influence their happiness. These themes are compared and contrasted to domains included in existing multi-dimensional measures of youth life satisfaction. The factors likely to be particularly salient to students with different levels of mental health are noted.

     

Qualitätsmanagement in organisierten HandlungssystemenQuality Management in Organized Systems of Action

Despite their prevalence and prominence, quality management in organizations as well as the various discourses on the topic lack theoretical framework and pervasion. The field is dominated by empirical-inductive experiences stemming from the so-called practice on the one hand and by pragmatic-normative suggestions for this practice. In contrast, for the purpose of a theoretically based analysis, we identify four implicit principles as the basal grammar of contemporary strategies of quality management, locate them within the Parsonian scheme of organized action systems, and embed them into the broader context of economy and society. This Parsonian perspective makes us aware of quality management as a one-dimensional imperative for action that in reality is still undermined, however, by subversive quality management practices, following a fifth principle.     

Retrofitting Drucker: Knowledge work under cognitive capitalism

This article uses Peter Drucker's work vector-like, to carry the thesis of cognitive capitalism into the management field. Drucker's prophetic insights into the knowledge society are juxtaposed with recent, Italian autonomist Marxist-inspired analyses of capitalism's cognitive phase. If the capacity to create knowledge – or what autonomists call the ‘general intellect’ – is becoming the key productive force, arguably there is a need for a full-blown social form of knowledge management. Our reading of Drucker thus retrieves the one worthwhile thing from the rubble of normative knowledge management. It is the idea of society-level knowledge management premised on a universal and unconditional guaranteed basic income (GBI; or social wage). Basic income represents not just a social investment in knowledge, which Drucker himself called for, but also compensation for biolabour's augmented social productivity. With Drucker as the steppingstone, we conclude, the autonomist tradition merits greater attention from critical management and organisational scholars interested in factoring class and gender back into the knowledge management equation.     

Critical Realism,Dialectics, and Qualitative Research Methods

Critical realism has been an important advance in social science methodology because it develops a qualitative theory of causality which avoids some of the pitfalls of empiricist theories of causality. But while there has been ample work exploring the relationship between critical realism and qualitative research methods there has been noticeably less work exploring the relationship between dialectical critical realism and qualitative research methods. This seems strange especially since the founder of the philosophy of critical realism, Roy Bhaskar, employs and develops a range of dialectical concepts in his later work in order to extend the main tenets of critical realism. The aim of this paper is to draw on Bhaskar's later work, as well as Marxism, to reorient a critical realist methodology towards a dialectical approach for qualitative research. In particular, the paper demonstrates how dialectical critical realism can begin to provide answers to three common criticisms made against original critical realist methodology: that the qualitative theory of causal powers and structures developed by critical realists is problematic; that critical realist methodology contains values which prove damaging to empirical research; and that critical realists often have difficulties in researching everyday qualitative dilemmas that people face in their daily lives.     

Generalizing the standardized hazard ratio to multivariate proportional hazards regression,with an application to clinical~genomic studies

The standardized hazard ratio for univariate proportional hazards regression is generalized as a scalar to multivariate proportional hazards regression. Estimators of the standardized log hazard ratio are developed, with corrections for bias and for regression to the mean in high-dimensional analyses. Tests of point and interval null hypotheses and confidence intervals are constructed. Cohort sampling study designs, commonly used in prospective–retrospective clinical genomic studies, are accommodated.     

Q-LEARNING WITH CENSORED DATA

We develop methodology for a multistage-decision problem with flexible number of stages in which the rewards are survival times that are subject to censoring. We present a novel Q-learning algorithm that is adjusted for censored data and allows a flexible number of stages. We provide finite sample bounds on the generalization error of the policy learned by the algorithm, and show that when the optimal Q-function belongs to the approximation space, the expected survival time for policies obtained by the algorithm converges to that of the optimal policy. We simulate a multistage clinical trial with flexible number of stages and apply the proposed censored-Q-learning algorithm to find individualized treatment regimens. The methodology presented in this paper has implications in the design of personalized medicine trials in cancer and in other life-threatening diseases.