管理支撑系统知识库

Previous studies focused on the characteristics of elicit scientists, or the early identification of high-potential scientists, but neglected career development research of ordinary young scientists. Using scientometrics and CV methods, the present study aims to establish statistic models in order to explore young scientists' career development, and predict their future academic impact. Predictors at 2009 were divided into four categories: publication indicators, scholar relationship indicators, scholar relationship indicators and research theme indicators, and these indicators were used to predict h-index at 2014. Predictive models were established separately in each category using stepwise regression, and finally a comprehensive model was built. Our result shows that hindex and total number of paper published could significantly predict scientific impact in the future. To improve academic impact, young scientists need to carry out extensive cooperation and play a key role in the cooperation, need to have the research experience in developed countries, and to reduce their research specialization.