跨学科流动对学者学术表现的影响———以人工智能领域为例

By constructing a database of academic performance and educational backgrounds of scholars in the field of artificial intelligence, and employing non - parametric tests, regression models, and other methods, this study examines the heterogeneous impact of multidisciplinary backgrounds on scholars' academic output. The findings reveal that multidisciplinary backgrounds influence scholars' academic performance, with specific differences observed in undergraduate and doctoral disciplines. Among doctoral graduates in computer science, electronic information, and automation, there are no significant differences in overall academic performance, but disparities exist in specific academic metrics such as the number of published papers. Furthermore, the undergraduate disciplines of these doctoral graduates do not significantly affect their future academic performance. Doctoral graduates in natural sciences exhibit slightly lower overall academic performance compared to those in computer science. However, among natural sciences doctoral graduates, those who studied automation as undergraduates perform significantly better academically than their peers with other undergraduate backgrounds. Similarly, humanities and social sciences doctoral graduates perform slightly worse overall than computer science doctoral graduates, but those with undergraduate backgrounds outside the humanities and social sciences outperform their peers with both undergraduate and doctoral degrees in the humanities and social sciences across all academic metrics. The conclusions offer insights for refining and adjusting interdisciplinary education policies in China, contributing to the transformation from a populous nation to a talent powerhouse.