中文摘要：

链式DEA模型研究科技投入产出两阶段效率时，忽视了阶段投入和成果产出之间的关联度，也没有就两阶段段效率对整体效率的促进作用进行检测，易造成投入因子和产出结果关联耦合度不足。在对各阶段的投入要素和成果产出进行灰色关联度分析的基础上，运用链式DEA模型对我国28个省市高技术产业的两阶段科技投入产出数据进行实证分析。然后借助单因子回归模型检验阶段效率对整体效率的作用。最后根据实证结果把各省市分为效率兼优型等四个类型。并针对性地提出政策建议。

英文摘要：

When using the chain DEA model to research the input and output of two-stage efficiency, the correlation between the stage input and output results is ignored, and two-stage segment efficiency role in promoting the overall efficiency of detection is not available, which could easily lead to low correlated coupling of factor inputs and outputs. Based on the grey relational analysis of various stages of input factors and output results, an empirical analysis on input-output data of high-tech industry＇s two-stage technology in China＇s 28 provinces is made by using the chain DEA model, and then the affection of the phase efficiency on the overall efficiency is verified by using the single-factor regression model testing. Finally, provinces and cities are divided into four types according to the empirical results, and policy suggestions are put forward.