Random effects regression mixtures for analyzing infant habituation

Random effects regression mixture models are a way to classify longitudinal data (or trajectories) having possibly varying lengths. The mixture structure of the traditional random effects regression mixture model arises through the distribution of the random regression coefficients, which is assumed to be a mixture of multivariate normals. An extension of this standard model is presented that accounts for various levels of heterogeneity among the trajectories, depending on their assumed error structure. A standard likelihood ratio test is presented for testing this error structure assumption. Full details of an expectation-conditional maximization algorithm for maximum likelihood estimation are also presented. This model is used to analyze data from an infant habituation experiment, where it is desirable to assess whether infants comprise different populations in terms of their habituation time.     

Still-births among the offspring of male radiation workers at the Sellafield nuclear reprocessing plant: detailed results and statistical aspects

Summary. This study investigates whether there was evidence of increasing risk of still-birth with increasing paternal exposure to ionizing radiation received during employment at the Sellafield nuclear installation before the child was conceived. A significant positive association is found between the total paternal preconceptional exposure to external ionizing radiation and the risk of still-birth (after adjustment for year of birth, social class, birth order and paternal age, odds ratio at 100 mSv 1.24 (95% confidence interval 1.04–1.45)). A summary of the principal scientific findings of this study has been published in the Lancet . This paper describes in detail the statistical methods that were used in the investigation and presents the results in full.     

Managing cross‐border cooperation between voluntary organizations in Ireland

The voluntary nonprofit sector in Ireland has grown significantly in recent years. A related trend has been the growth of cross‐border (in effect, transnational) cooperation between voluntary organizations based in Northern Ireland and the Republic of Ireland. However, this development has posed a set of management issues in terms of structure, forms of governance and decision making, day‐to‐day project management, and sustainability. Cross‐border work has also meant that voluntary bodies confront a distinctive set of barriers in relation to political and ethnic conflict, social and cultural attitudes, and practical delivery. Strengthening of cooperation across borders will require greater planning, mainstream funding, focused management, diversification of services, and realistic expectations.     

Knowledge,power, and ethics in qualitative social research

Philadelphia, then and still one of the worst governed of America’s badly governed cities, was having one of its periodic spasms of reform. A thorough study of causes was called for. Not but what the underlining cause was evident to most white Philadelphians: the corrupt semi-criminal vote of the Negro Seventh Ward…. Would it not be well, to elucidate the known causes by a scientific investigation?… With my bride of three months, I settled in one room over a cafeteria run by a College Settlement, in the worst part of the Seventh Ward. We lived there a year, in the midst of an atmosphere of dirt, drunkenness, poverty and crime. Murder sat on our doorstep, police were our government, and philanthropy dropped in with periodic advice. (DuBois, 1968/1975: 194–5 also in Logan, ed. 1944: 42) What had been borne into me during my book studies was my utter ignorance of the manual-working class, that is, of four-fifths of my countrymen…. How was I to get an opportunity of watching, day by day, in their homes and in their workshops, a sufficient number of normal manual-working families to enable me to visualize the class as a whole; to understand what was meant by chronic poverty and insecurity of livelihood. (Webb, 1926: 146–7)     

The calibration of P-values,posterior Bayes factors and the AIC from the posterior distribution of the likelihood

The posterior distribution of the likelihood is used to interpret the evidential meaning of P-values, posterior Bayes factors and Akaike's information criterion when comparing point null hypotheses with composite alternatives. Asymptotic arguments lead to simple re-calibrations of these criteria in terms of posterior tail probabilities of the likelihood ratio. (Prior) Bayes factors cannot be calibrated in this way as they are model-specific.