科学生产过程的数学模型和中国大学科研能力  

A Stochastic Model for Scientific Production Processes & the Distributive Pattern of Scientific Productivity on Chinese Campus

作　　者：山石[1] 龚礼明[1] 凌柳玉[1] 

机构地区：[1]上海大学,上海市200072

出　　处：《科学学研究》2000年第1期19-28,共10页Studies in Science of Science

摘　　要：本文利用随机过程构造了一个数学模型来刻划具有“成功产生成功”效应的科学生产过程。该模型能够精确地预测科学生产率的数量分布 ;采用二种新方法推导了科学生产率的理论分布———韦林分布 ;给出了韦林分布的数字特征 ,还提出了估计参数的二种新方法 ;转换数据法 (TDE)和极大似然法 (MLE) ;为了验证该数学模型的正确性 ,给出了 1991- 1995年共五年科学生产率在中国各大学分布的观察值和理论值。A study on the distributive pattern of scientific productivity is presented in the article from methodological and mathematical angles. Rooted on the success-breeds-success effect, the processes of scientific production are modelled and evolved into a stochastic process and through two ways derived from the model, the Waring distrbution is developed. In consideration of the parameters in the Waring distribution to be estimated, two estimators-the transformed data estimator(TDE) and the maximum likelihood estimator(MLE)-are evaluated respectively. In addition, properties owned by the Waring distribution are discussed one by one. In order to verify the model, an empirical study on the distributive layout of scientific productivity on Chinese campus is provided in the article.

关 键 词：科学计量学 科学生产过程 科学生产率 韦林分布 中国 大学 科研能力 

分 类 号：G301[文化科学] G649.2