Bayesian analysis of outlier problems using divergence measures

A Bayesian approach is presented for detecting influential observations using general divergence measures on the posterior distributions. A sampling-based approach using a Gibbs or Metropolis-within-Gibbs method is used to compute the posterior divergence measures. Four specific measures are proposed, which convey the effects of a single observation or covariate on the posterior. The technique is applied to a generalized linear model with binary response data, an overdispersed model and a nonlinear model. An asymptotic approximation using Laplace method to obtain the posterior divergence is also briefly discussed.     

Two point priors and Γ-minimax estimating in families of uniform distributions

It is shown that when a parameter lying in a sufficiently small interval is to be estimated in a family of uniform distributions, a two point prior is least favourable under squared error loss. The unique Bayes estimator with respect to this prior is minimax. The Γ-minimax estimator is derived for sets Γ of priors consisting of all priors that give fixed probabilities to two specified subintervals of the parameter space if a two point prior is least favourable in Γ.     

A probablistic analysis for an optimal screening problem

Summary We consider a lotL formed byN apparently similar unitsW ₁,…,W _N, where each of theW _i may come from one of two different populationsP ₁ andP ₂;T ₁,…,T _N denote the corresponding lifetimes. The units fromP _i undergo a failure of kindi and their survival function isS _i (t). We assume that the failure rate function are known and that the units fromP ₁ are ?substandard?: λ ₁ (t)≥λ ₂ (t), ∀t≥0. We want to putW ₁,…,W _N under a pre-operational test (burn-in test) in order to eliminate at least a great part of the substandard units and we face the problem of obtaining a rule for stopping the test under the assumption that, with the failure of a unit, it is possible to recognize the population from which the unit comes. Such a problem will be formalized as an optimal stopping problem for a suitably defined Markov process. Our study shall evidentiate some fundamental aspects of the problem and the role of the prior distribution of the (random) numberM ₀ of those units inL coming fromP ₁ (substandard). The latter distribution has a great influence on the form of the solution. This research was supported by the C.N.R. Project ?Statistica Bayesiana e Simulazione in Affidalità e Modellistica Biologica?.     

The Evolution of Knowledge and the Dynamics of an Industry Network

The paper moves a step forward in the direction of establishing a connection between the structure and evolution of knowledge bases and the structure and evolution of organizational forms in innovative activities in a science-intensive industry. The paper has an explicit focus on the dynamics of the network of collaborative agreements in R&D in the pharma/biotech industry after the “molecular biology revolution”. Using a comprehensive dataset, built by the authors integrating several sources in the industry, the dynamics of the network over time is extensively analyzed. With regards to network structure, it is found that, while the size of the network increases over time due to net flows of entry, its topological properties remain relatively unchanged. The evolution of the network has occurred without relevant deformations in the core-periphery profile. With regards to age-dependent propensity to collaborate, the paper finds that the extent of inter-generational collaboration is much more significant than intra-generational collaboration. In addition, the propensity of firms of a given generation to enter into collaboration with firms of a different generation increases with the distance between the two, while the total number of intra-generational collaborations decreases over time and, moreover, tends to decrease for most recent generations. In the paper a unitary and coherent explanation of the evidence is developed, coming to reveal the existence of a striking isomorphism between structural properties of the dynamics of knowledge and of the evolution of network structure. This revised version was published online in July 2006 with corrections to the Cover Date.     

Innovation through initiatives—a framework for building new capabilities in public sector research organizations

The accelerating pace of change in science and technology has resulted in new attention to the process of identifying and developing ideas that ultimately lead to new scientific capabilities and business opportunities for an organization. The need to refresh research programs and capabilities is as important in federally funded research institutions as it is for industry. This paper explores the critical success factors for new initiatives at a federal laboratory, and building on lessons learned through this study and in private industry, identifies a more systematic process that could potentially improve the effectiveness of these initiatives in achieving results.     

债券增信对企业创新的影响研究

债券增信能帮助企业以更好的条件发行债券,因而对企业的创新能力产生影响。通过对2007-2015年上市债券的研究发现：债券增信能够显著提升企业创新能力,但不同的增信方式（第三方保证、抵押和质押）其影响存在差异,其中,以采用抵押担保的增信方式效果为最佳;并且,债券增信对企业创新的提升效果在融资约束企业更为明显,从另一侧面证明了融资约束对企业创新的负面影响。     

高管薪酬激励与企业创新——基于中国上市公司的经验证据

创新是经济增长的重要驱动力,在当前我国由“中国制造”迈向“中国智造”的关键阶段,企业要获取更多的利润,必须依靠创新来强化自身的核心竞争力。高管是企业的关键决策者,对企业创新活动具有重要影响。文章利用2010—2017年中国949家上市公司的面板数据,首先,分别采用面板固定效应模型和面板负二项回归模型实证检验了高管薪酬激励对企业研发投入和研发产出的影响,研究发现,高管薪酬激励的三种模式,即货币薪酬、股权和在职消费激励均显著促进了企业创新活动。其次,文章还探讨了高管薪酬激励对企业创新活动的异质性影响。一方面,根据企业所有权性质将样本企业划分为国有和非国有企业样本组,研究发现,与国有企业相比,高管薪酬激励显著促进了非国有企业的创新活动。另一方面,根据企业所属行业类别将样本企业划分为高新技术企业与非高新技术企业样本组,研究发现,股权激励模式显著提高了高新技术企业的研发投入和非高新技术企业的研发产出。为检验上述实证结果的可靠性,文章还采用更换企业研发投入和研发产出指标的衡量方法进行稳健性检验,结果显示,高管薪酬激励与企业研发创新的显著正相关关系依然稳健。高管自身风险厌恶程度和个人能力等因素不仅会...     

中间品贸易视角下的国际研发合作行为研究

随着全球化程度加深,与国外企业进行研发合作逐步成为国内企业参与创新驱动发展战略的重要途径.从中间品贸易视角出发,通过构建谈判博弈模型,分析探讨影响产业链下游企业进行国际研发合作内生性选择的因素.研究发现,只有在研发溢出效应比较小的时候,产业链下游企业才会选择进行国际研发合作.但如果研发溢出效应过小,下游企业的国际研发合作行为会导致本国社会福利水平下降.进一步研究发现,上游市场垄断势力增强以及最终品贸易成本下降均会对产业链下游企业的国际研发合作产生促进作用.     

Analysis of child labour in Peru and Pakistan: A comparative study

This paper analyses child labour participation and its key determinants using data sets from Peru and Pakistan. The results include tests of the ‘Luxury’ and ‘Substitution’ hypotheses that play key roles in recent studies on child labour and child schooling. The results reject both hypotheses in the context of child labour in Pakistan and suggest that income and related variables do not have the expected negative effect on children's work input. Rising wages of adult female labour in Pakistan, and falling adult male wage in Peru lead to increased participation of children in the labour market. The results on the combined country data formally establish the presence of strong individual country effects in the estimated regressions. For example, ceteris paribus, a Peruvian child is more likely to experience schooling than a Pakistani child. However, both countries agree on the positive role that adult female education and infrastructure investment in basic amenities can play in discouraging child labour and encouraging child schooling. Received: 24 August 1998/Accepted: 10 March 1999