Practical Issues in Implementing and Understanding Bayesian Ideal Point Estimation

Andrew Gelman Department of Statistics and Department of Political Science, Columbia University, New York, NY e-mail: gelman{at}stat.columbia.edu, www.stat.columbia.edu/~gelman/ David K. Park Department of Political Science, Washington University, St. Louis, MO e-mail: dpark{at}artsci.wustl.edu Noah Kaplan Department of Political Science, University of Houston, Houston, TX e-mail: nkaplan{at}uh.edu Logistic regression models have been used in political sciencefor estimating ideal points of legislators and Supreme Courtjustices. These models present estimation and identifiabilitychallenges, such as improper variance estimates, scale and translationinvariance, reflection invariance, and issues with outliers.We address these issues using Bayesian hierarchical modeling,linear transformations, informative regression predictors, andexplicit modeling for outliers. In addition, we explore newways to usefully display inferences and check model fit.     

The developmentally handicapped witness

This study compared the ability of 30 developmentally handicapped (DH) and 30 nonhandicapped individuals in their ability to report on witnessed events. Participants watched a film clip and were asked to respond to five different types of questions about it. Misleading or leading questions were embedded in three of the question types Results indicated that in response to free recall and very general questions, DH participants did not provide as complete responses as did nonhandicapped participants, although the information provided by both groups tended to be quite accurate. Both groups, and particularly DH individuals, were less accurate in responding to more focused shortanswer recall questions. In response to correct leading specific and statement questions, requiring recognition of the information provided, DH and control participants were comparable in their performance. DH participants had greater difficulty with misleading recall questions and false leading specific and statement questions. Results are discussed in terms of optimizing eyewitness accuracy in DH individuals.     

Correction to: Beyond Performance: Racial Prejudice and Whites’ Mistrust of Government

Political Behavior -     

INCORPORATING ATTACHMENT ASSESSMENT INTO CUSTODY EVALUATIONS: THE CASE OF A 2‐YEAR‐OLD AND HER PARENTS

Attachment with parents is central to a child's development. It is well established that the quality of this attachment in early childhood is a strong predictor of developmental and psychological functioning throughout the life span. One of the primary issues in custody evaluations is assessing the quality of the child's attachment to each parent and the parents' capacity to foster security and to consider what this might mean for short‐ and medium‐term decisions about their care. The nature of attachment measures is summarized, and the combined use of three attachment‐caregiving instruments in a custody evaluation is illustrated through the case of a toddler whose parents were engaged in a high‐conflict divorce. The case study demonstrates how, in addition to standard clinical observations, including a set of attachment‐based instruments with a standardized psychological test battery provided information critical to a recommendation for custody and parent visitation.     

Assessing nation-state-sponsored cyberattacks using aspects of Situational Crime Prevention

Research Summary

This study utilized a quantitative analysis of 246 cyberattack incidents reported in the Extremist CyberCrime Database to identify significant predictors of nation-state-sponsored cyberattacks relative to those performed by non-nation-state-sponsored ideological actors. Clarke and Newman's Situational Crime Prevention framework for terrorism was used to identify differential opportunities to successfully affect targets on the basis of tools, weapons, and the ability to access targets in online settings. The analysis noted nation-state-sponsored attacks were less likely to use high-visibility attack methods and more likely to utilize attack methods leading to data breaches. In addition, they were more likely to target state governments and military entities relative to ideological actors.

Policy Implications

Nation-state attacks are more difficult to identify or mitigate while in process, requiring a more robust national cybersecurity policy framework to be implemented that moves beyond current practices. There is a need to better utilize all aspects of government, from legislation to grant funding, in order to deter cyberattacks from continuing into the future.