Calibrated path sampling and stepwise bridge sampling

A computational problem in many fields is to evaluate multiple integrals and expectations simultaneously. Consider probability distributions with unnormalized density functions indexed by parameters on a 2-dimensional grid, and assume that samples are simulated from distributions on a subgrid. Examples of such unnormalized density functions include the observed-data likelihoods in the presence of missing data and the prior times the likelihood in Bayesian inference. There are various methods using a single sample only or multiple samples jointly to compute each integral. Path sampling seems a compromise, using samples along a 1-dimensional path to compute each integral. However, different choices of the path lead to different estimators, which should ideally be identical. We propose calibrated estimators by the method of control variates to exploit such constraints for variance reduction. We also propose biquadratic interpolation to approximate integrals with parameters outside the subgrid, consistently with the calibrated estimators on the subgrid. These methods can be extended to compute differences of expectations through an auxiliary identity for path sampling. Furthermore, we develop stepwise bridge-sampling methods in parallel but complementary to path sampling. In three simulation studies, the proposed methods lead to substantially reduced mean squared errors compared with existing methods.     

碳管制效率、政府能力与碳排放

 在碳约束时代追求经济增长与环境保护双赢发展的要求下,探索兼顾发展与控制的平衡路径具有重要的意义。本文构造了一个引入随机边界分析的STIRPAT模型,从国际比较的视角,研究政府能力和碳管制效率对控制碳排放的影响作用。研究结论显示,过去10多年来,各国的碳管制效率并没有太大的改进,且在2008年金融危机期间还出现了明显的下降,表明各国在探索提升碳管制效率以及有效控制碳排放方面还有很长的路要走。中国的碳管制效率总体上位居世界前列,超过了美国和欧盟,但低于日本,碳管制领域的法制力度和政府效率是影响中国碳管制效率提升的主要因素,而加强政府能力将有助于提升碳管制效率,进而促进节能减排。     

An evaluation in the choice of inputs and outputs in the efficiency measurement of police forces

The new Labour government has recently instigated an initiative to establish whether English and Welsh police forces should be ranked into groups based on an efficiency measure. A particular problem, however, is that policing includes many inputs and outputs (outcomes) that could potentially be utilised in efficiency models. This paper considers the problems associated with measuring relative police force efficiency given that a vast number of potential indicators (inputs and outputs) must be reduced to a handful to allow feasible estimation. In addition, it discusses the input and output variables utilised in the first ‘official’ analysis of English and Welsh police force efficiency [Demonstration Project, Home Office, London, UK, 2001].     

Just-in-time in ATO vehicle manufacturing: The Minivan-Piaggio V.E. case study

The system adopted by Piaggio V.E. to run the production of 'Minivan' aims at conciliating the manufacturing philosophy 'just-in-time' JIT with the creation of a wide range of end-product codes. Some general assumptions for a JIT production, with small lot sizes, are described together with their implementation in the operationorganizing system adopted by Piaggio V.E. for the 'Minivan' assembly line. The strategy pursued by Piaggio V.E. for the Minivan focuses on the quality of the product and the level of service in the form of fast order cycle and a high number of options offered on catalogue. According to the logic of an ATO production system, the customer orders a customized product and the manufacturer does not keep any finished inventory, but aims at effective management of the information flows.     

One-sided EWMA control charts for monitoring Poisson processes with varying sample sizes

ABSTRACT

Recently considerable research has been devoted to monitoring increases of incidence rate of adverse rare events. This paper extends some one-sided upper exponentially weighted moving average (EWMA) control charts from monitoring normal means to monitoring Poisson rate when sample sizes are varying over time. The approximated average run length bounds are derived for these EWMA-type charts and compared with the EWMA chart previously studied. Extensive simulations have been conducted to compare the performance of these EWMA-type charts. An illustrative example is given.     

Interactive effects of perceived control and job stressors on affective reactions and health outcomes for clerical workers

Karasek's (1979) hypothesis that perceived control interacts with various job stressors in affecting employee satisfaction and health was tested. It was proposed that high levels of perceived stress would only be associated with poor health and negative affect in the presence of low control. One hundred and thirty-six clerical workers at a major US university completed questionnaries containing the measures of interest. The results of regression analyses failed to support the interaction hypothesis. However, measures related to both control and job stressors were found to correlate with satisfaction and health outcomes, as has been found in prior research. Limitations of the self-report and correlational methodology are discussed.     

Optimization of design parameters for control charts with multiple assignable causes

Duncan's economic model of Shewhart's original x¯ chart has established its optimal and economic application for processes with the Markovian failure characteristic. As the sample statistics show some indications of process variations, the variable-sampling-interval (VSI) control charts perform more effectively than the fixed sampling interval (FSI) ones due to a higher frequency in the sampling rate. Regarding the economic design of control charts, most studies have been dedicated to the FSI scheme. In 1998, Bai & Lee considered the production process with a single assignable cause and proposed an economic VSI design for a general x¯ control chart. However, in real cases, there are multiple assignable causes in the production process. Therefore, concerning the operation characteristics of the real industry, this research develops an economic model for the VSI control chart with multiple assignable causes based on stochastic and statistics theory and determines the optimal design parameters of the chart. A numerical example is also provided to demonstrate the effectiveness of the proposed model and the result indicates that VSI performs more effectively than a FSI control chart.     

“We Actually Trust the Community:” Examining the Dynamics of a Nonprofit Funding Relationship in New Zealand

Funding relationships in nonprofit management are increasingly defined by a philosophy of rational management, characterized by measurement of outputs and benchmarking, which represents an audit culture system (Burnley, Matthews, & McKenzie, 2005). There is concern that these approaches are constantly undermining the mission of community service nonprofit organizations (Darcy, 2002). In this research, we analyzed the management of funding relationships by examining dynamics within a nonprofit funding relationship in New Zealand. Through focus groups we explored the relationship between 17 representatives from nonprofit organizations and four Board members of a funding Trust. The management of this funding relationship was characterized by an appreciation of the diverse nature of nonprofit organizations, a balance between trust and control, and communication. We suggest that elements of these dynamics could be incorporated into nonprofit funding relationships in order to challenge an over-reliance on audit culture systems, and to re-establish relationships characterized by interaction between nonprofit organizations and their funders. Finally, we call for future research in this area.     

我国上市公司的财务困境成本及其影响因素分析

本文选择1998—2002年间沪深股市40家ST摘帽公司A股为样本,从“经营业绩观”和“权益价值观”两个角度考察我国上市公司是否存在财务困境成本。结果发现:(1)从陷入财务困境之前到解除财务困境之后,基于“经营业绩观”,企业经行业调整之后的平均主营业绩有所增长,但基于“权益价值观”,经过市场调整之后的权益市场价值平均下降了2.04%,表明投资者平均承担2.04%的财务困境成本,说明虽然企业在解除财务困境之后经营业绩较之于陷入财务困境之前有所增长,但这种经营业绩上的增长并未带来权益市场价值的提高。(2)行业因素和企业规模是影响财务困境的重要因素。当财务困境企业所在行业业绩不佳时财务困境成本较高;企业的资产规模与财务困境成本呈显著的正向关系。     

The use of Andrews curves for detecting the out-of-control variables when a multivariate control chart signals

The aim of this paper is to present a new method for solving the problem of detecting the out-of-control variables when a multivariate control chart signals. The main idea is based on Andrews curves. The proposed method is investigated thoroughly and is proved to have interesting results in comparison to a competing method.