信息隐私作为一个长期存在的问题，成为影响个人乃至整个国家安全的重要议题。其中，“隐私管理”是我国在社交媒体时代探索隐私保护的重要切入点，而本研究将通过平台差异和个体特质这两个方面对社交媒体用户的隐私管理行为（包括边界联系性、边界渗透性和边界所有权）进行探究。具体而言，第3章基于多维隐私发展理论探讨了隐私认知的概念内涵。首先，本研究对38位被试者进行深度访谈，确定了隐私关注和隐私经验的两个内在结构，并编制了一个初步量表。为检验量表，本研究发放了三次问卷（N1=316；N2=525；N3=732）。研究结果显示，该量表具有较好的信效度。并且，两个维度和25个测试题项的隐私认知量表被得以验证。第4章采用潜在剖面分析，以自我汇报的方式通过问卷调查法（N=1127）来确定不同社交媒体用户的隐私认知类型，并进一步考察了不同隐私认知类型与人口统计学变量之间的关系。结果发现，我国社交媒体用户的隐私认知类型（从认知程度高到低）包括“隐私基要主义者”“隐私实用主义者”“一般隐私关心者”和“隐私漠不关心者”四种类型。此外，不同隐私认知用户类型在年龄和收入方面存在显著差异，而在性别、受教育水平、使用社交媒体的年限方面不存在显著差异。第5章在沟通隐私管理理论和可供性理论的基础上，从微信、微博和小红书平台的角度考察了隐私认知、隐私可供性和边界管理（边界联系性、边界所有权、边界渗透性）之间的关系。结果显示，隐私认知对边界所有权与边界渗透性具有正向预测作用；隐私认知对边界联系性无显著预测作用。针对微信这类强关系平台，隐私认知通过隐私可供性间接提升边界联系性和边界所有权；针对微博和小红书这类弱关系平台，隐私认知通过隐私可供性间接提升边界联系性和边界渗透性。第6章检验了隐私认知、隐私犬儒主义、实际隐私素养和边界管理之间的关系。研究发现，隐私关注与边界联系性和边界渗透性之间呈负相关，而与边界所有权呈正相关。隐私经验和边界所有权之间呈现显著正相关，而与边界渗透性和边界联系性之间未呈现显著相关关系。此外，隐私犬儒主义与边界联系性和边界渗透性呈正相关，而与边界所有权呈负相关。最后，隐私经验会通过实际隐私素养对边界联系性产生作用；实际隐私素养会在隐私犬儒主义和边界联系性之间发挥中介作用。
Information privacy, as a long-standing issue, has become a crucial concern affecting individual and even national security. Privacy management is an important channel for China to explore privacy protection in the era of social media. This study focuses on privacy cognition and boundary management, investigating the relationships among privacy cognition, privacy affordance, real privacy literacy, and boundary management from the perspectives of platform differences and individual traits. Specifically, in the first sub-study, the concept of privacy cognition is explored based on the Multidimensional Privacy Development Theory (MPDT). This study conducted in-depth interviews with 38 participants to identify two underlying structures of privacy concerns and privacy experience, leading to the development of a preliminary scale. To validate the scale, three surveys were conducted (N1=316; N2=525; N3=732), and the scale was tested using pre-test, exploratory factor analysis (EFA), and confirmatory factor analysis (CFA). This results indicate good reliability and validity of the scale, validating the privacy cognition scale with two dimensions and 25 test items. In the second sub-study, latent profile analysis was employed to determine different types of privacy cognition among social media users through self-reported questionnaires (N=1127). Additionally, the study explored the relationships between different privacy cognition types and demographic variables. The findings revealed that four types of privacy cognition among Chinese social media users, ranging from high to low cognitive levels: “privacy fundamentalists,” “privacy pragmatists,” “general privacy enthusiasts,” and “privacy indifferent individuals.” Furthermore, significant differences were observed in age and income among different privacy cognition types, while no significant differences were found in gender, education level, and the duration of social media usage. The third sub-study, based on Communication Privacy Management (CPM) theory and affordance theory, examined the relationships among privacy cognition, privacy affordance, and boundary management (boundary linkage, boundary ownership, boundary permeability) from the perspectives of different platforms (WeChat, Weibo, and Xiaohongshu). The results indicated that privacy cognition positively predicts boundary ownership and boundary permeability, while having no significant predictive effect on boundary linkage. Specifically, for the WeChat platform, privacy cognition indirectly enhances boundary linkage and boundary ownership through privacy affordance; for Weibo and Xiaohongshu platforms, privacy cognition indirectly improves boundary linkage and boundary permeability through privacy affordance. The fourth sub-study tested the relationships among privacy cognition, privacy cynicism, real privacy literacy, and boundary management. The research revealed a negative correlation between privacy concerns and boundary linkage and boundary permeability, with a positive correlation with boundary ownership. Additionally, significant positive correlations were found between privacy experience and boundary ownership, while no significant correlations were observed between privacy experience and boundary permeability or boundary linkage. Furthermore, privacy cynicism showed positive correlations with boundary linkage and boundary permeability, and a negative correlation with boundary ownership. Finally, privacy experience had an effect on boundary linkage through real privacy literacy, and real privacy literacy played an intermediate role between privacy cynicism and boundary linkage.
