【摘要】 不同于相对容易量化的科学研究评价，学科声誉评价多集中于客观数据上的主观评价，产生新的偏差，带有偏差的结果会直接影响知识生产方式、教师学生等利益相关者的行为、价值观和社会学想象力。为此，探究学科声誉评价新方法，以工商管理学科为例，采集5848名首次入职教职岗位的博士毕业生数据，建立从就业单位到博士学位授予单位的有向邻接矩阵，运用社会网络分析的理论和方法对博士互聘网络进行量化分析。结果表明：(1)博士互聘网络能反映学科点的学缘偏好结构，是度量学科声誉的新指标；(2)从高校间聘用行为方向看，头部高校间存在水平流动及向下流动，中部及以下高校间以向下流动为主。高校人才招聘与其学术成果产出具有滞后性，因此，对基于博士毕业生“用脚投票”的博士互聘网络结构特征和网络演化趋势的挖掘，比基于后续成果指标的学科排名能更快反映学科质量的变化，可用来对学科质量变化进行预警。更多还原

【Abstract】 Unlike scientific research evaluation which is relatively easy to quantify, discipline reputation evaluation is mostly based on subjective evaluation of objective data, resulting in new deviations, which will directly affect the behavior, values and social imagination of stakeholders such as knowledge production methods, teachers and students. In order to solve problems above and explore a new method of discipline reputation evaluation, it take the business administration discipline as an example, collect the data of 5848 PhD graduates who first entered teaching posts, establish a directed adjacency matrix from the employment unit to the doctoral degree awarding unit, and use the theory and method of social network analysis to conduct quantitative analysis on the PhD exchange network. The results show that:(1) PhD exchange network can reflect the academic preference structure of doctoral program, and is a new indicator to measure discipline reputation;(2) From the perspective of employment behavior among universities, there is horizontal flow and downward flow between the head universities, and downward flow is mainly among the middle and lower universities. There is a certain time lag between the recruitment of university talents and the output of their academic achievements. Therefore, the mining of the structural characteristics and network evolution trend of the PhD exchange network based on the “foot voting” of PhD graduates is faster than the discipline ranking based on the follow-up achievement indicators to reflect the changes in the discipline quality, which can be used to warn of changes in the discipline quality.更多还原