This paper optimized the e-commerce live streaming atmospheric cues scale and then used the SOR model to study the impact of atmospheric cues on consumers' impulse buying behaviors while considering the mediating role of social presence. Questionnaires were collected from consumers with e-commerce live streaming shopping experience, and Mplus 7.4 was utilized for analyzing the measurement model and structural models. We optimized the existing atmospheric cues scales by adding interactive cues, music design, and other elements to better measure the atmospheric cues in e-commerce live streaming. The findings confirm the reliability and validity of the e-commerce live streaming atmospheric cues scale, comprised of four dimensions: information, navigation, design, and interactive cues. Additionally, information, design, and interactive cues were positively associated with consumers' impulse buying behaviors, while the mediating role of social presence partially explains these relationships. However, the relationship of navigation cues on impulse buying behaviors is not significant. This optimized scale provides a valuable measurement tool for measuring atmospheric cues in future research. Furthermore, the empirical study sheds light on the antecedents of impulse buying behaviors by examining the impact of atmospheric cues and the mediating role of social presence. The results offer practical suggestions for e-commerce live streaming operators to increase consumer purchases: (1) Optimize the design cues by carefully designing colors and layout, appropriately using and coordinating colors and lighting, and playing suitable background music. (2) Provide diversified and personalized information by enhancing the professionalism of the broadcasters and promptly addressing viewer questions from the bullet chat. (3) Actively engage consumers by inviting them to participate in the live streaming sessions, creating an immersive shopping experience.

本文对电商直播氛围线索量表进行优化，然后在考虑社交临场感的中介作用的情况下，利用SOR模型研究氛围线索对消费者冲动购买行为的影响。对有电商直播购物经历的消费者进行问卷调查，利用Mplus 7.4分析测量模型和结构模型。我们对现有的氛围线索量表进行了优化，加入了互动线索、音乐设计等元素，以更好地衡量电商直播中的氛围线索。研究结果证实了电商直播氛围线索量表的信度和效度，该量表由信息、导航、设计和互动线索四个维度组成。此外，信息、设计和互动线索与消费者的冲动购买行为呈正相关，而社交存在的中介作用部分解释了这些关系。然而，导航线索与冲动购买行为的关系并不显着。这种优化的尺度为未来研究中测量大气线索提供了有价值的测量工具。此外，实证研究通过考察大气线索的影响和社会存在的中介作用，揭示了冲动购买行为的前因。研究结果为电商直播运营商增加消费者购买提供了切实可行的建议：（1）优化设计线索，精心设计色彩和布局，适当使用和协调色彩和灯光，播放合适的背景音乐。（2）通过提升主播专业水平，及时解答观众弹幕提问，提供多元化、个性化的信息。（3）邀请消费者参与直播，积极吸引消费者，打造沉浸式购物体验。