CONDITIONAL INFERENCE ABOUT AN EQUICORRELATION COEFFICIENT

SenGupta (1987) proposed a locally most powerful test which is globally (one sided) unbiased, and an estimator of p, the equicorrelation coefficient of a standard symmetric multivariate normal (SSMN) distribution. Here we use the idea in Williams (1984) to illustrate the construction and use of ancillary statistics to make inference about p. The test and confidence intervals based on this construction are conditionally optimal.     

The Age Difference Between Spouses: Variations among Developing Countries

Sub-national and cross-national variations in the age difference between spouses are investigated with data from the World Fertility Survey relating to 29 developing countries. Substantial variation within and between countries is evident. Analysis suggests that the relative age of prospective spouses is a factor taken into account in the marriage market. Observed variation in the age difference, within and between countries, cannot be explained as the simple by-product of the random matching of independently determined distributions of men's and women's ages at marriage. Certain age differences are avoided, others chosen more frequently. Preferred age differences appear to differ in the societies studied, however, and this variation can be directly interpreted in terms of two sets of factors: kinship structure and women's roles. The analysis also suggests that demographic determinants of the age difference, in particular age constraints on the pool of possible matches, are of less importance in explaining societal variations than are social structural factors.     

Playground Accessibility and Neighbourhood Social Interaction Among Parents

While the positive association between social interaction and access to green space is well accepted, little research has sought to understand the role of children??s playgrounds in facilitating social interaction within a community. Playgrounds are spaces designed to facilitate play and the interaction of children, but may also be important places of interaction between parents. In this paper we examine how access to playground spaces is related to social interaction between parents. We use two measures of accessibility (1) walking distance to the closest playground and (2) playground service area, a measure of the number of potential users of a playground based on population density. We use generalized estimating equations, an extension of generalized linear models, to control for the confounding effects of socio-economic status (income, education), neighbourhood dynamics (neighbourhood location, years in neighbourhood) and free time (daily outdoor activity, marital status, number of children) on the independent relationship between social interaction and access to playground spaces. Our results suggest that while accessibility to playgrounds is associated with social interaction among parents, the direction of the effect is opposite to existing literature on green space and social interaction; parents with low accessibility to playgrounds are more likely to interact socially with their neighbours than parents with high accessibility. Our results suggest a pattern of spatial behaviour in which the burden of poor access to some resources may actually encourage greater neighbourhood engagement. Future research studying the relationship between health and green space may benefit from studying the specific role of playground spaces.     

t-Latinized Designs

A variety trial sometimes requires a resolvable block design in which the replicates are set out next to each other. The long blocks running through the replicates are then of interest. A t -latinized design is one in which groups of these t long blocks are binary. In this paper examples of such designs are given. It is shown that the algorithm described by John & Whitaker (1993) can be used to construct designs with high average efficiency factors. Upper bounds on these efficiency factors are also derived.     

Forecasting mortality rates via density ratio modeling

The authors propose a semiparametric approach to modeling and forecasting age‐specific mortality in the United States. Their method is based on an extension of a class of semiparametric models to time series. It combines information from several time series and estimates the predictive distribution conditional on past data. The conditional expectation, which is the most commonly used predictor in practice, is the first moment of this distribution. The authors compare their method to that of Lee and Carter.     

Knowledge, Consequences, and Experience: The Social Construction of Environmental Problems

Perhaps one of the most obvious yet difficult questions confronting sociologists concerned with large-scale environmental problems is an epistemological one: How do we know what we know about the state of the environment? This paper explores the realist and constructionist approaches to environmental-social problems and finds both inadequate as currently formulated. A case is made for a phenomenological constructionism that moves beyond relativism and simple definitional constructionism by exploring how we actually experience the world. This approach recognizes the existence of a natural world independent of our constructions, yet suggests that our knowledge of it is always mediated, indirect, and pragmatically motivated.     

Minimum wage effects by gender

Recent criticisms have led some to dismiss time-series analyses in the debate over the minimum wage. We investigate previous time-series studies showing that raising the minimum wage has a smaller impact on females than males. We reanalyze the data in light of recent developments in time-series methods and find that the minimum wage has a similar significant negative impact on both males and females. We conclude that, following a 10 percent increase in the minimum wage, both male and female employment drops from between 2 and 4 percent over a two-year period. This employment decrease slowly erodes as economic growth and inflation cause the minimum wage to fall below the market-clearing wage. We thank David Card and Alan Krueger for generously providing the data.     

Changing the Palace Guard: Analysing the Impact of Women's Entry into Medicine

Women's rapid entry into medicine raises important questions about change in this historically male-dominated profession. In addition to shifting the gender balance, do increasing numbers of women signal a more fundamental change in the way in which medicine is organized and practised? A growing body of research suggests such change. Yet there is continuing controversy about the meaning of observed gender differences in practice, whether they reflect an essentially pragmatic response to women's dual workload of family and career, or whether they are the result of underlying differences in attitudes, values and orientations. This article uses data from a survey of a recent cohort of family physicians in Ontario, Canada, to document the extent of gender differences and similarities in medical practice, and their interrelationships to family situation, political attitudes and patient care attitudes. The results show that seemingly related differences in practice are accounted for in different ways: while some are associated with differences in family situation, others are tied to attitudes, while others are a function of gender alone. The results also reveal gender similarities which do not suggest that women are becoming more like men, but that men, and the profession as a whole are changing. We would emphasize the importance of interpreting gender differences and similarities within a broader conceptual understanding of change in the profession.     

Mood and Audience Effects on Video Lottery Terminal Gambling

Little is known about the situational factors associated with gambling behavior. We induced 180 male participants (mean age: 21.6) into a positive, negative, or neutral mood prior to gambling on a video lottery terminal (VLT). While gambling, participants were observed by either a male peer, female peer, or no one. Induced mood had no effect on gambling behavior. Participants induced into a negative mood prior to gambling, however, reported more positive moods after gambling, whereas those with positive and neutral moods reported more negative moods after gambling. Participants observed by either a male or female peer spent less time gambling on the VLT compared to those not observed. Participants observed by a female peer lost less money relative to the other observer conditions. Degree of problem gambling in the last year had little influence on these effects. Some practical implications of these findings are discussed.