Stochastic blockmodels: First steps

A stochastic model is proposed for social networks in which the actors in a network are partitioned into subgroups called blocks. The model provides a stochastic generalization of the blockmodel. Estimation techniques are developed for the special case of a single relation social network, with blocks specified a priori. An extension of the model allows for tendencies toward reciprocation of ties beyond those explained by the partition. The extended model provides a one degree-of-freedom test of the model. A numerical example from the social network literature is used to illustrate the methods.     

Suicide ideation,attempts and abuse among incarcerated gang and nongang delinquents

The present study examines suicide ideation and attempts as well as reported levels of physical and sexual abuse among gang and nongang-involved incarcerated youth. The self-report survey data were collected from 334 males and 61 females who were incarcerated in Nevada youth correctional facilities in the summer of 1994. Over half (58.3 percent) of the males and nearly half (45.2 percent) of the females reported being gang members. Surprisingly high rates of abuse (more than 70 percent for both physical and sexual abuse) were found among the incarcerated females of this study. Females reported significantly more suicide attempts and were more likely to have been physically and sexually abused than males. Nongang males reported significantly more suicide attempts and were more likely to report sexual abuse than gang-involved males. Among the gang-involved delinquents, females reported significantly more suicide attempts than males. In addition, gang members who reported sexual abuse were more likely to have higher levels of suicide ideationand to have attempted suicide. Implications for youth correction specialists and social workers are discussed.The data for this study were collected under Grant #000616 from the Nevada Agricultural Experiment Station.     

Bayes and empirical Bayes shrinkage estimation of regression coefficients

For a moderate or large number of regression coefficients, shrinkage estimates towards an overall mean are obtained by Bayes and empirical Bayes methods. For a special case, the Bayes and empirical Bayes shrinking weights are shown to be asymptotically equivalent as the amount of shrinkage goes to zero. Based on comparisons between Bayes and empirical Bayes solutions, a modification of the empirical Bayes shrinking weights designed to guard against unreasonable overshrinking is suggested. A numerical example is given.     

Enhancing the chances of successful OR/MS implementation: the role of the advocate

This paper extends understanding of the role of the advocate in the implementation of Operations Research/Management Science models in governmental agencies. The first part of this paper presents and analyzes the data resulting from interviews with 39 criminal justice agencies regarding the model implementation process. The interview data indicated that the advocate performed important problem-solving and mediation functions needed to introduce the technology. The second part of this paper provides a case study describing an advocate's successful introduction of a Patrol Car Allocation Model (PCAM) in the Los Angeles County Sheriff's Department. To assess the overall effectiveness of their implementation strategy, the project team conducted a questionnaire attitudinal survey of lieutenants, sergeants and officers in two of the pilot stations prior to PCAM implementation and after staffing changes. Overall, the attitude of the respondents was more favorable at the time of the second survey. The analysis of the interviews and the PCAM case study suggests a simple normative model for enhancing the chances of successful implementation of Operations Research/Management Science models.