Bayesian multiple comparison of models for binary data with inequality constraints

In this paper we consider generalized linear models for binary data subject to inequality constraints on the regression coefficients, and propose a simple and efficient Bayesian method for parameter estimation and model selection by using Markov chain Monte Carlo (MCMC). In implementing MCMC, we introduce appropriate latent variables and use a simple approximation of a link function, to resolve computational difficulties and obtain convenient forms for full conditional posterior densities of elements of parameters. Bayes factors are computed via the Savage-Dickey density ratios and the method of Oh (Comput. Stat. Data Anal. 29:411–427, 1999), for which posterior samples from the full model with no degenerate parameter and the full conditional posterior densities of elements are needed. Since it uses one set of posterior samples from the full model for any model in consideration, it performs simultaneous comparison of all possible models and is very efficient compared with other model selection methods which require one to fit all candidate models. A simulation study shows that significant improvements can be made by taking the constraints into account. Real data on purchase intention of a product subject to order constraints is analyzed by using the proposed method. The analysis results show that there exist some price changes which significantly affect the consumer behavior. The results also show the importance of simultaneous comparison of models rather than separate pairwise comparisons of models since the latter may yield misleading results from ignoring possible correlations between parameters.     

Going Places: Effects of Early U.S. Compulsory Schooling Laws on Internal Migration

Population Research and Policy Review - Both the industrialization thesis and institutional theories of education hypothesize that early educational expansion increased internal migration. We take...     

A hybrid genetic algorithm for train sequencing in the Korean railway

This article addresses the train-sequencing problem encountered in the Korean railway. It first presents a mixed integer programming model for the problem, in which the mileage must be balanced for each train route, while various field constraints must be satisfied, including overnight stay capacity and maintenance allocation restrictions. Then, it proposes a hybrid genetic algorithm as a solution approach to the problem. The proposed algorithm utilizes a modified elite group technique along with two heuristic procedures based on the mixed integer programming model. Finally, the proposed solution approach is tested with real-world data from the Korean railway. Numerical experiments under different conditions indicate that the proposed solution approach to the train-sequencing problem is promising.     

Kernel partial correlation: a novel approach to capturing conditional independence in graphical models for noisy data

Graphical models capture the conditional independence structure among random variables via existence of edges among vertices. One way of inferring a graph is to identify zero partial correlation coefficients, which is an effective way of finding conditional independence under a multivariate Gaussian setting. For more general settings, we propose kernel partial correlation which extends partial correlation with a combination of two kernel methods. First, a nonparametric function estimation is employed to remove effects from other variables, and then the dependence between remaining random components is assessed through a nonparametric association measure. The proposed approach is not only flexible but also robust under high levels of noise owing to the robustness of the nonparametric approaches.     

Mildly explosive autoregression with mixing innovations

In this paper, the limit distribution of the least squares estimator for mildly explosive autoregressive models with strong mixing innovations is established, which is shown to be Cauchy as in the iid case. The result is applied to identify the onset and the end of an explosive period of an econometric time series. Simulations and data analysis are also conducted to demonstrate the usefulness of the result.     

Family Social Networks and Recovery From Severe Mental Illness of Clubhouse Members

This study examined the contribution of family social network variables to recovery among Clubhouse members. A random sample of 118 members from one Clubhouse was recruited for this cross-sectional study. Among all respondents, greater overall family support and better relationship quality with the most supportive network member were correlated with greater levels of recovery. Clubhouse members perceived higher quantity and quality of support and more criticalness from their most supportive network members, principally family members, compared to the rest of their family members. Longitudinal studies are needed to further investigate the relationship between family support and recovery.     

Generation and Earnings Patterns Among Chinese,Filipino, and Korean Americans in New York

By treating the 1.5 generation as a distinctive analytic category, this paper compares the effects of generational status on earnings among men of Chinese, Filipinos, and Korean descents in the New York metropolitan area. Our analyses of the 5 percent Public Use Microdata Sample data of the 2000 U.S. census show that all other background characteristics held equal, 1.5‐generation Chinese and Filipino American workers make significantly higher earnings than second‐generation workers. However, Korean American workers do not exhibit this 1.5‐generation advantage. These findings support a segmented assimilation theory, the view that immigrant assimilation paths are not uniform across ethnic groups or generation status. Other findings suggest that bilingual ability would increase earnings only for the Chinese group.     

Why the Home Region Matters: Location and Regional Multinationals

Much of the literature in international business analysing the multinational enterprise uses the country as the relevant environmental parameter. This paper presents both theoretical and empirical evidence to demonstrate that country‐level analysis now needs to be augmented by analysis at the ‘regional’ level of the broad triad markets of Europe, North America and the Asia Pacific. The great majority of the world's 500 largest firms concentrate their activities within their home region of the triad. This study uses variance component analysis and finds that this home region effect outperforms the country effect. Together, the regional and industry effects explain most of the geographic expansion of multinational enterprises (MNEs), whereas country, firm and year effects are very minor. The new data and variance component analysis on the activities of large MNEs reported here suggest that new thinking is required about the importance of large regions of the triad as the relevant unit of analysis for business strategy to supplement the conventional focus on the country.     

A descriptive study of some organizational-behavior dimensions at work in the Dominican Republic: implications for management development and training

Based in identity theory and the depletion hypothesis in the work-life literature, this study investigated dimensions of the meaning of work and work stress among mid-level managers in the United States, Brazil, and Korea. Findings include a high salience of work as a central life concern but a primacy of family involvement in each country. The importance of religious involvement, leisure activities, and community involvement was rated low. Intrinsic and extrinsic orientations to working were related to increased levels of work stress, but in each country the dimensions of work stress were associated with meaning of work aspects in different combinations and affected different demographic groups in different ways. The theoretical and practical implications of the findings imply the need for a differentiated understanding of the meaning of work dimensions in international settings and the awareness that increased levels of work stress can results from higher levels of work centrality and intrinsic and economic work orientations.     

Bayesian analysis of time series Poisson data

This paper provides a practical simulation-based Bayesian analysis of parameter-driven models for time series Poisson data with the AR(1) latent process. The posterior distribution is simulated by a Gibbs sampling algorithm. Full conditional posterior distributions of unknown variables in the model are given in convenient forms for the Gibbs sampling algorithm. The case with missing observations is also discussed. The methods are applied to real polio data from 1970 to 1983.