Variance partitioning in multilevel logistic models that exhibit overdispersion

Summary.  A common application of multilevel models is to apportion the variance in the response according to the different levels of the data. Whereas partitioning variances is straightforward in models with a continuous response variable with a normal error distribution at each level, the extension of this partitioning to models with binary responses or to proportions or counts is less obvious. We describe methodology due to Goldstein and co-workers for apportioning variance that is attributable to higher levels in multilevel binomial logistic models. This partitioning they referred to as the variance partition coefficient. We consider extending the variance partition coefficient concept to data sets when the response is a proportion and where the binomial assumption may not be appropriate owing to overdispersion in the response variable. Using the literacy data from the 1991 Indian census we estimate simple and complex variance partition coefficients at multiple levels of geography in models with significant overdispersion and thereby establish the relative importance of different geographic levels that influence educational disparities in India.     

Identifying abused children using assessments and observations in the classroom: a preliminary study

The paper presents two pilot studies that attempt to identify maltreated children in the classroom. The aim of pilot study 1 was to determine whether six British physically abused children performed signi?cantly worse than matched non‐abused children on a battery of tests and rating scales over an 18‐month period. The aim of pilot study 2 was to concurrently identify maltreated children from a classroom of children using standardized tests and observed negative behaviours. The method used in study 1 was a cross‐comparison design with six physically abused and 12 comparison children, all between 4 and 8 years old. They were assessed over 18 months at four time‐points on a battery of standardized tests and rating scales. Study 2 was based on a concurrent prediction of child abuse in a class of 30 children (18 male and 12 female) 10 years old. They were assessed on a battery of standardized assessments and by the direct observation of negative behaviour in order to predict those who were on the child protection register (CPR) for actual or suspected abuse and those who were not. The results of study 1 showed signi?cant differences on the teacher rating scales and on a measure of sibling dependency. Study 2 used standardized assessments and identi?ed three of ?ve abused children. Direct observations identi?ed one of ?ve abused children. These two preliminary studies suggest standardized assessments are more useful than behavioural observations in identifying abused children in the classroom setting. Further con?rmation is required using larger‐scale investigations. Copyright © 2003 John Wiley & Sons, Ltd.     

Beyond rural idylls: Imperfect lesbian utopias at Michigan Womyn’s music festival

Whilst rural idylls have dominated some discussions of rural social difference, little attention has been paid to rural utopias. Imagined, material and discursive experiences of utopian rural ideals are critically examined in this paper. It takes as its focus the Michigan Womyn’s Music Festival - an annual US womyn-only festival - in order to understand the mutual construction of empowering rural spaces and non(hetero)normative women’s bodies. The first part of the paper moves away from the current focus on ‘rural idylls’ to understand social difference and ruralities, and argues that the notion of ‘rural utopia’ is a useful framework to reconsider rural possibilities for those considered ‘other’ to dominant (heterosexual) hegemonies. The context of the festival and qualitative empirical research methodologies are explained. In the second part of the paper four themes that arose from this research are explored. Firstly, the ‘free expression’ of sexualities, including sex and relationships is seen as creating a ‘slice of lesbian rural utopia’ at Michfest. Secondly, rural utopian understandings of Michfest as womyn’s space where diverse bodily aesthetics are celebrated are compared and contrasted to urban spaces, and particularly the gendered embodiments of ‘everyday’ life. Third, while womyn’s bodies may contest heteronormative performances of femininity, they also reiterate normative and embodied versions of ‘women’, creating womyn’s space through the controversial ‘womyn-born-womyn’ entry policy. Fourth, rural tensions are examined around the negotiation of work (voluntary and paid). The paper concludes by arguing that a focus on imperfect (lesbian) utopias furthers our understanding of gender, sexuality and rurality.     

Developing a clinical care pathway for obese pregnant women: A quality improvement project

ProblemObesity in pregnancy is associated with an increased incidence of maternal and foetal morbidity and mortality, from conditions like preeclampsia, gestational diabetes, preterm birth and stillbirth. Between 20% and 25% of pregnant women in Australia are presenting to their first antenatal appointment with a body mass index (BMI) ≥30 kg/m², defined as obesity in pregnancy. These figures are concerning for midwifery and obstetric staff directly involved in the clinical care of these women and their families. In the absence of national or state clinical practice guidelines for managing the risks for obese pregnant women, a local quality improvement project was conducted.AimTo plan, implement, and evaluate the impact of an alternative clinical care pathway for pregnant women with a BMI ≥ 35 kg/m² at their first antenatal visit.Project settingThe project was undertaken in the antenatal clinic of a rural referral hospital in NSW, Australia.SubjectsEighty-two women with a BMI ≥ 35 kg/m² were eligible for the alternative care pathway, offered between January and December 2010.InterventionThe alternative care pathway included the following options, in addition to usual care: written information on obesity in pregnancy, referral to a dietitian, early plus repeat screening for gestational diabetes, liver and renal function pathology tests, serial self-weighing, serial foetal growth ultrasounds, and a pre-labour anaesthetic consultation.FindingsDespite being educated on the risk associated with obesity in pregnancy, women did not take up the offers of dietetic support or self-weighing at each antenatal visit. Ultrasounds were well received and most women underwent gestational diabetes screening.     

Embracing the “And”: Between Queer and Bisexual Theory at Brighton BiFest

Questions of bi identities can be invisibilized and overlooked by queer theorizing and LGBT studies. This article explores the ways in which complex performances of bisexuality can simultaneously encompass and deconstructively critique bi identity in a manner that embraces the “and” between bi and queer, offering important insights into how bi is lived, contested, and reaffirmed. Drawing on the BiCon and BiFest events in the UK, we argue that both the materialities (and supposed fixities) of bi erasures and exclusions and the fluidities that trouble the heterosexual/homosexual divides offer key insights into the spatial and temporal fixing and unfixing of identities.     

A population-level metric for gambling-related harm

Prior estimates of the population-level impact of gambling have relied on economic costings. Recent work has derived disability weights for the Problem Gambling Severity Index, which measure per-person impact of gambling on quality of life on a scale of zero to one. This provided scope for the present study to calculate the ‘burden of gambling harm’, which captures the aggregate impact of harms arising from gambling on quality of life in a population. Gambling-related harm was associated with 101,675 years of life lost in Victoria, Australia: approximately two-thirds that of alcohol use and dependence, and major depressive disorder. Problem gamblers suffer more individually (disability weight = .44) compared to those in moderate (.29) and low (.13) risk categories. Nevertheless, moderate and low-risk gamblers account for 85% of population-level harm, due to greater prevalence of these groups. Overall, the scale of gambling-related harm is large relative to other significant health issues, with milder yet non-negligible harm accruing to a relatively broad segment of the gambling population. The article suggests that the tendency to conflate the (typically low) prevalence of problem gambling with total gambling impact is misleading, and argues for a broader population-health based measure.     

A typical problem gambler affects six others

While the financial and psychological burden on problem gamblers can be severe, at least some of the ill effects are also passed on to family or other close social ties. The present study estimated the number of affected-others for the typical problem gambler. Australian members of an online panel with gambling problems (N = 3076) and panel members who indicated that they had been affected by someone else’s gambling (N = 2129) were asked to estimate the number of other people who were negatively affected by their gambling. Using robust statistics to analyse this data, the study found lower estimates made by problem gamblers (four affected people) compared to estimates made by affected others (six affected people, including the respondent). It was concluded that a point-estimate of six people affected is a more accurate figure since it does not suffer from self-presentation effects of problem gamblers. Low-risk and moderate-risk gamblers, unsurprisingly, affected far fewer other people (one and three, respectively). Both gamblers and affected-others most often identified close family members, including spouses and children, as the people impacted by others’ gambling problems. These results provide an approximate measure of the number of people affected, per problem gambler, to facilitate accurate accounting of the harms accruing from gambling problems.