Abstract: Virtual doctor team is a novel model on the online health consultation platform. Incorporating information technology, it can provide more timely and comprehensive consultation services for patients. It plays an important role in reasonably matching doctors and patients and promoting hierarchical diagnosis and treatment. However, there are few studies on the mechanism of virtual doctor teams. This paper incorporates an Input-Process-Output (IPO) theoretical model to investigate how knowledge diversity and task dependence among team doctors affects their virtual team performances and how the frequency of interaction among team members moderates the effects of knowledge diversity and task dependence on team performance. In this study, the panel data of eight consecutive months of a virtual doctor team from the Good Doctor Online Website were collected, and panel negative binomial regression random effect model was used to verify those hypothesis. The empirical results show that knowledge diversity and task dependence have a significant positive impact on team performance, and contact frequency among members negatively moderates the impact of task dependence on team performance. This conclusion broadens the theoretical application of virtual doctor teams and provides practical guidance for the establishment of efficient virtual doctor teams on online health consultation platform.

Key words: online doctor team, knowledge diversity, task dependence, contact frequency, team performance

摘要： 医生团队模式是在线健康咨询平台的一个新兴模式，借助信息技术，可以为患者提供更及时、更全面的咨询服务，对于合理匹配医生患者并促进分级诊疗有重要作用。本文运用了团队绩效研究中的理论模型“投入-过程-输出”（Input-Process-Output，IPO）研究了虚拟医生团队成员间的知识多样性、任务依赖性如何影响团队绩效，并检验了团队成员的互动过程接触频率如何调节知识多样性、任务依赖性与医生团队绩效的关系。通过网络爬虫共收集了好大夫网站上连续8个月的面板数据, 并使用面板负二项回归随机效应模型验证假设。实证结果显示，知识多样性和任务依赖性都显著正向影响团队绩效，接触频率负向调节任务依赖性对团队绩效的影响。研究结论拓宽了虚拟团队的理论应用，也为在线健康咨询领域提供了组建高效医生团队的实践指导。

关键词: 在线医生团队, 知识多样性, 任务依赖性, 接触频率, 团队绩效