本文运用三阶段DEA模型，对2014年我国区域科普投入产出效率进行了研究。在剔除了人口受教育程度、科技发展水平、网络化环境等三个环境变量和统计噪音的影响后，有2/3的省份在技术效率或者纯技术效率上有效，这是一个比较可喜的现象。但全国平均规模效率值下降较大，13个规模报酬递增的省份需扩大规模。东部和西部地区尤需提高规模效率在科普资源优化配置中的作用，中部地区需在科普管理、技术水平上向先进地区追赶。

In this paper, a three-stage DEA model is used to study the input-output efficiency of regional science popularization of China in 2014. Two thirds of provinces are effective in terms of technical efficiency or pure technical efficiency after excluding three environmental variables and statistical noise, such as the level of education, the level of technological development, and the network environment, which is a relatively promising finding. However，the national average scale efficiency decreased significantly, and 13 provinces with increasing returns to scale need to expand the scale of science popularization. For the eastern and western regions, it is necessary to improve the scale efficiency in the optimal allocation of science popularization resources. The central region is expected to catch up with the advanced areas in the management and technical level of science popularization.

G315

科技部科技基础性工作专项“全国科普资源调查与数据分析基础性工作”（2014FY130600）