Statistical Inference and Forensic Evidence: Evaluating a Bullet Lead Match

This experiment tested the ability of undergraduate mock jurors (N=295) to draw appropriate conclusions from statistical data on the diagnostic value of forensic evidence. Jurors read a summary of a homicide trial in which the key evidence was a bullet lead "match" that was either highly diagnostic, non-diagnostic, or of unknown diagnostic value. There was also a control condition in which the forensic "match" was not presented. The results indicate that jurors as a group used the statistics appropriately to distinguish diagnostic from non-diagnostic forensic evidence, giving considerable weight to the former and little or no weight to the latter. However, this effect was attributable to responses of a subset of jurors who expressed confidence in their ability to use statistical data. Jurors who lacked confidence in their statistical ability failed to distinguish highly diagnostic from non-diagnostic forensic evidence; they gave no weight to the forensic evidence regardless of its diagnostic value. Confident jurors also gave more weight to evidence of unknown diagnostic value. Theoretical and legal implications are discussed.     

Correlates of correctional officer job stress: The impact of organizational structure

Corrections can be a very stressful occupation, producing many negative outcomes for employees. These outcomes, both mental and physical, not only affect the employee, but the entire organization. A large body of literature has focused on identifying correlates of correctional staff job stress; yet, not all of the possible correlates have been studied. One area that has received little, if any, attention is the impact of the organizational structure on correctional job stress. Organizational structure includes issues of centralization, instrumental communication, integration, and organizational justice. The results of multivariate analysis indicate that instrumental communication and procedural justice have a direct effect on correctional staff job stress.     

The economic impact of engineering research centers: preliminary results of a pilot study

This article describes the results of a pilot study that tested the feasibility of estimating quantitatively the regional and economic impacts of NSF-supported Engineering Research centers. For regional impacts, we combined estimates of the direct plus indirect and induced economic impacts of ERC expenditures generated from a regional input–output model with estimates of the additional impact on the state due to center-based start-up companies, licensing income from intellectual property produced by the center, the cost savings enjoyed by local firms that had hired center graduates, and advice and consulting to local firms by center faculty. For national economic impact, a suitably modified version of the regional approach was employed, supplemented by use of a consumer surplus model to estimate the net public benefits of newly commercialized technologies based in center research. As the project proceeded, it became clear that efforts to focus solely on economic impacts that could be quantified relatively easily would greatly underestimate the actual national economic impact of ERCs. The types of impacts included and the kinds of data collected from centers and their collaborating companies were therefore expanded in the later case studies. Results of the first three cases are described here; findings from the remaining two studies did not change our overall results or conclusions. The profile of regional and, especially, national economic impact estimates varied widely across the centers studied. Only some of these variations could be attributed to ERC characteristics; most were the result of variations in the amount and type of data that could be obtained from the centers involved and the companies they worked with. We concluded that even the most conscientious and costly data collection efforts would be unlikely to yield comparable data across centers because the accessibility of key data, especially proprietary data, will differ unpredictably from center to center. Further, focusing on narrowly-conceived, quantifiable economic data alone should be avoided in these kinds of impact studies. Doing so distorts the amount and characteristics of actual impacts, many of which—perhaps most of which—cannot feasibly be converted to monetary terms. Such a narrow focus will greatly underestimate the impact of ERC-like centers, masking the much broader and, based on our findings, larger and more significant impacts on society.     

Public administration planning in developing countries a Bayesian decision theory approach

Decisionmaking under uncertainty is visualized as a game against nature. The policymaker is the player and has a set of alternatives or strategies from which he desires to choose the most effective. He is confronted with a variety of possible states of nature that may evolve after his decision is made. The key states of nature are identified and their interactive relationship to the strategy options are specified. Important roles in the resulting analysis are played by the probabilities of successful project initiation and implementation. It is shown that neglect of the inherent uncertainty aspects leads to evaluations (benefit-cost ratios) of proposed reforms and projects which can be seriously upward biased.     

美国大学的通识教育

生活是一个整体,而非划分为不同的专业。现代社会中的工作、生活都已超越了传统的专业边界,所有的学生都需要通识教育以获得全面发展。美国德克萨斯A&M大学设立的校级核心课程、文学院课程和国际研究课程,对其他大学开展通识教育可能会有参考作用。     

A comparative cost analysis of commodity foods from the U.S. Department of Agriculture in the National School Lunch Program

Schools that participate in the National School Lunch Program receive a portion of their federal funding as commodity foods rather than cash payments. This research compared the product costs and estimated total procurement costs of commodity and commercial foods from the school district perspective using data from 579 Minnesota ordering sites in school year (SY) 2008–2009. Though comparison of product prices indicates that commercial foods were an average of 17 percent more expensive than equivalent commodity foods, once full estimated procurement costs were included, the commercial products became 9 percent less expensive per food case than commodity products. Base case results were sensitive to the cost of risk to hold inventory. When this input was tested at zero, the estimated cost to procure commercial foods was 2 percent higher per case than commodity foods, though total commercial procurement costs remained less than commodity costs by 3 percent. It is estimated that Minnesota schools collectively spent an additional $1.7 to $3.7 million to procure USDA commodity foods in SY 2008–2009. © 2009 by the Association for Public Policy Analysis and Management.