Predictors of High-Risk Sexual Behavior Among Individuals Living with HIV Disease and AIDS

An important focus of the 1996 World Aids Conference related to the extent of unprotected sex among gay men, particularly those in their twenties. This paper will use logistic analysis to explore the extent of unsafe sexual practices among individuals living in Bexar County, Texas, who are HIV+ assymptomatic, HIV+ symptomatic, or have AIDS. Using survey research data, we assess the significance and impact of various independent factors on whether respondents reported they did or did not engage in unprotected sex. The most striking finding is that unprotected sexual activity is prevalent regardless of demographic and/or status differences among respondents. Findings reinforce the use of community planning processes, which emphasize the importance of sound behavioral and social science combined with efforts to include the affected constituency to create effective HIV prevention programs.     

Bayesian Inference Under Partial Prior Information

Partial prior information on the marginal distribution of an observable random variable is considered. When this information is incorporated into the statistical analysis of an assumed parametric model, the posterior inference is typically non‐robust so that no inferential conclusion is obtained. To overcome this difficulty a method based on the standard default prior associated to the model and an intrinsic procedure is proposed. Posterior robustness of the resulting inferences is analysed and some illustrative examples are provided.     

Gender Role Ideology and the Gender Based Differences in Earnings

Much of the research on gender differences in occupational earnings still focuses on human capital and the structure of the labor market. However, these variables rarely explain even half of the gender gap in earnings. Most research has examined the impact of gender role ideology as it impacts occupational choice, which indirectly can impact earnings. Using data from the National Opinion Research Center General Social Surveys, this research focuses on the relationship between attitudes about gender roles and two variables: (a) earnings, and (b) occupational positions held by women and men. Findings show that traditional gender-role ideology contributes to lower observed earnings for both males and females, independent of the influences of human capital characteristics, occupational context, and ascribed characteristics. Results support socialization as a partial explanation for the gender-based earnings differences and suggest that, to the extent that economic rewards are used to assess the value of gender role expectations, traditional gender role attitudes might continue to change and lead to relatively equal earnings among women and men.     

Sexist ideology and the evaluation criteria used to assess women's integration into the army

This article is a reaction to the 1978 Evaluation of Women in the Army report issued by the U.S. Army. It looks at the criteria used in women's evaluation in terms of the sexist ideology prevalent in the military system. Those issues which are identified by the Army as problematic and which affect policies regarding women's full integration into the Army are discussed at length.An earlier version of this paper was presented at the annual meeting of the Southern Sociological Society in April, 1983. I am indebted to Nijole V. Benokraitis for her comments at that time. I would also like to thank John Sibley Butler, Susan Marshall and Marie Crane for their helpful comments. Any opinions, conclusions, and recommendations, whether expressed or implied, are those of the author.     

On regression-based tests for persistence in logarithmic volatility models

Building on the work of Pantula (1986), this paper discusses how the hypothesis of conditional variance nonstationarity in the logarithmic family of generalized autoregressive conditional heteroskedasticity (GARCH) and stochastic volatility processes may be tested using regression-based tests. The latter are easy to implement, have well-defined large-sample distributions, and are less sensitive to structural changes than tests based on the quasimaximum likelihood estimator.     

A Bayesian method of distinguishing unit root from stationary processes based on panel data models with cross-sectional dependence

In this paper we develop a Bayesian approach to detecting unit roots in autoregressive panel data models. Our method is based on the comparison of stationary autoregressive models with and without individual deterministic trends, to their counterpart models with a unit autoregressive root. This is done under cross-sectional dependence among the error terms of the panel units. Simulation experiments are conducted with the aim to assess the performance of the suggested inferential procedure, as well as to investigate if the Bayesian model comparison approach can distinguish unit root models from stationary autoregressive models under cross-sectional dependence. The approach is applied to real exchange rate series for a panel of the G7 countries and to a panel of US nominal interest rates data.