[目的/意义] 旨在构建适用于中文科研合作成果的跨学科性测量方法，并系统分析科研合作组织特征对知识成果跨学科性的影响规律。该研究有助于弥补现有测量方法主要面向英文文献、在中文情境下适用性不足的缺陷，拓展跨学科测量的本土化研究工具。同时，可为优化科研合作组织方式、完善跨学科成果评价和合理配置科研资源提供参考。[方法/过程] 提出一种基于中图分类号的中文科研成果跨学科性测量方法，能够应用于大规模文献数据集，相较基于参考文献的测量，该方法能够很好地排除内生性影响。以2013—2022年62家CSSCI期刊的124 746篇合作论文为样本，采用含年份固定效应与期刊固定效应的Tobit模型进行实证检验。[结果/结论] 研究发现，在控制年份与期刊固定效应后，基金支持数量、内部组织规模与外部合作规模均对知识成果跨学科性产生显著倒“U”型影响，表明适度的合作规模有助于跨学科知识整合，但超过一定阈值后协调成本上升将产生抑制效应。上述结论在去极端值处理和替代口径检验中均保持稳健，对科研合作的组织管理具有重要实践启示。研究表明，科研合作治理应由单纯的规模扩张导向转向更加注重协同质量和知识整合效率的质量导向。

[Purpose/Significance] This paper aims to develop a method for measuring the interdisciplinary nature of Chinese scientific research collaborations and to systematically analyze the patterns by which the characteristics of research collaboration organizations influence the interdisciplinary nature of knowledge outputs. Accurate measurement of interdisciplinarity is a prerequisite for evaluating collaborative research quality, optimizing research resource allocation, and advancing interdisciplinary research policies. A framework tailored to Chinese-language literature also helps address the limited applicability of existing approaches, which have primarily been developed for English-language databases. [Method/Process] The study proposed a method for measuring the interdisciplinary nature of Chinese research outputs based on the Chinese Library Classification (CLC) system, which was applicable to large-scale literature datasets and helped mitigate potential endogeneity concerns compared to citation-based measurement methods. Specifically, CLC codes were mapped to disciplinary categories at different levels to identify disciplinary combinations and calculate the interdisciplinarity of individual papers. Using a sample of 124 746 collaborative papers published in 62 CSSCI journals between 2013 and 2022, the study conducted empirical tests with a Tobit model incorporating year and journal fixed effects. [Result/Conclusion] The study finds that, after controlling for year and journal fixed effects, the number of funding grants, internal team size, and the scale of external collaboration all exert significant inverted U-shaped effects on the interdisciplinary nature of research outputs. At moderate levels, larger teams, broader institutional participation, and increased funding support expand access to heterogeneous knowledge, skills, and resources, thereby promoting knowledge recombination. However, once collaboration exceeds an appropriate threshold, rising communication, coordination, and organizational costs inhibit interdisciplinary knowledge integration. These findings remain robust after removing extreme values and using alternative measurement specifications. The proposed method provides a feasible tool for measuring the interdisciplinarity of large-scale Chinese research outputs, while the empirical findings offer practical guidance for configuring research teams, coordinating interorganizational collaboration, and allocating funding support.