随着数字技术的深度渗透,企业在追求效率和竞争力的同时也面临着数据滥用、算法歧视等数字伦理风险,因此企业数字责任逐渐进入理论研究视野。但是,企业数字责任现有研究尚存在概念界定模糊、维度结构分散、应用测量薄弱等问题,制约着理论探索的进一步深化。本研究旨在构建企业数字责任的结构体系和测度方案。首先,在回顾现有企业数字责任概念研究基础上,立足数字化核心构建了企业数字责任结构框架,其中企业社会责任居于表层进行整体价值引领,数字要素、数字技术、数字管理和数字价值遵循“起点—过程—治理—结果”的动态逻辑构筑底层内核,承载和实现数字责任;其次,通过构建特征词库、自然语言处理、企业粉饰处理和数据校正等程序和方法设计了企业数字责任测度方案;最后,选择汽车制造行业开展实际测算,初步验证了该测度方案的有效性。本研究有助于增强企业数字责任概念的清晰性和可测性,不仅为企业数字责任定量研究奠定了变量与测度基础,也为推广企业数字责任评估提供了指标与方案借鉴。

With the deep penetration of digital technologies, while pursuing efficiency and competitiveness, enterprises are also confronted with digital ethical risks such as data misuse and algorithmic discrimination. Consequently, corporate digital responsibility (CDR) has increasingly attracted scholarly attention. However, extant studies on CDR still face challenges such as ambiguous conceptual definitions, fragmented dimensional structures, and weak measurement approaches, which constrain further theoretical exploration. This study aims to construct a structural framework and measurement scheme for CDR. First, building on a review of extant conceptual research, this study develops a digitalization-centered structural framework of CDR. In this framework, corporate social responsibility serves as an overarching value anchor at the upper level, while digital elements, digital technologies, digital management, and digital value constitute the core architecture following a dynamic "starting point-process-governance-outcomes" logic, through which CDR is articulated and operationalized. Second, a measurement scheme for CDR is designed by integrating procedures and methods including the construction of a feature lexicon, natural language processing, corporate impression management correction, and data calibration. Finally, the automotive manufacturing industry is selected for empirical measurement, providing initial evidence for the effectiveness of the proposed scheme. This study contributes to enhancing the clarity and measurability of CDR concept, not only laying a foundational for variable construction and measurement in quantitative research, but also providing referential indicators and a feasible scheme for assessing CDR.