Abstract: With the rapid development of information technology, online innovation communities have become an important way for many companies to obtain user innovation resources to promote open innovation. However, many communities still face the problem of low activity and insufficient user contributions. In recent years, scholars and managers have paid great importance to the driving factors of users’ contribution behavior. In online communities, online communication expands the scope of people’s interaction, and user frequent interactions build a complex social network, which also forms the most essential difference between online communities and face-to-face user communication. So, under such conditions, how does user interaction in social networks affect users’ contributions behavior? This research will answer this question. Based on the Stimulus-Organism-Response(S-O-R) theory, a conceptual model is proposed to explain how peer influence (breadth, depth) affect users’ contribution behavior (proactive contribution, responsive contribution), and the mediating effect of perceived benefits between peer influence and users’ contribution behavior. The research takes the creative market section of Xiaomi community MIUI function and discussion area as the object, a total of 6567 user data were collected through python, including 8830 ideas and 5,2560 comments posted by users, as well as 103,3558 comments received and so on, and use Mplus8.1 to analyze and test the data. A structural equation model, regression analysis, and bootstrapping method were applied to analyze the objective data from the network of Xiaomi. The research results show that: (1) peer influence (breadth and depth) positively affect users’ contribution behavior; (2) user’s integrative benefits positive mediates the relationship between peer influence and users’ contribution behavior; (3) user’s affective benefits positive mediates the relationship between peer influence and proactive contribution behavior; (4) user’s cognitive benefits negative mediates the relationship between peer influence depth and responsive contribution behavior. Therefore, this research has important theoretical and practical significance. In theory, previous research on peer influence is common in the field of sociology, such as research on effect of adolescent’s antisocial behavior. The online community creates a network-based user social environment that promotes user interaction. The paper introduces peer influence into the study of user behavior in online innovation communities, and deepens the research on the effect of peer influence on users’ contribution behavior in online innovation communities. Previous studies mainly examined the influence of users’ internal and external motivation factors, external environmental factors on users’ contribution behavior, ignoring the social network context of online communities. At the same time, we expand the literature on peer influence in social networks, supplementing the literature on peer influence on users’ contribution behavior in online communities. Second, we aim to expand the understanding of the connotation of peer influence in online communities. We divide the peer influence into the peer influence breadth and peer influence depth, and deeply explore the influence of the horizontal and vertical changes in the user’s knowledge base brought about by peer influence in the community on the user’s contribution behavior, it makes up for the singularity and limitation of the dimensionality of peer influence in previous studies. Third, we apply S-O-R theoretical model in the field of environmental psychology to the field of online user innovation research, and introduces three types of perceived benefits from the user interaction in online innovation communities: cognitive, affective and integrative benefits, and different from prior studies on the benefits with self-reported survey data , we capture the user’s perceived benefits in objective archival comments made by users, deeply discussion of the mediate effect of perceived benefits between peer influence and users’ contribution behavior. In a practical sense, in order to encourage users to actively participate in knowledge contribution, community managers should pay attention to the interaction between community members, and actively encourage community members to share their ideas on online community platforms to stimulate other users to produce convergence behavior. In addition, the community should strengthen the “social design” of online community platforms, and strengthen social functions and features, which will help users’ initial adoption, continuous participation and user retention, thereby creating and maintaining a platform that helps users obtain more perceived benefits.

摘要： 本文将同侪影响引入在线创新社区的用户行为研究中，从广度和深度两方面考察同侪影响对用户贡献行为的影响，并分析感知收益的中介作用。研究以小米社区MIUI功能与讨论区的创意集市板块为对象构建S-O-R模型，采用6567名用户发布的8830条创意、5.26万条评论和收到的103.36万条评论数据，利用Mplus8.1分析检验，结果发现：同侪影响广度与深度均有利于促进用户贡献行为，综合收益在同侪影响广度、深度与用户贡献行为间起正向中介效应，情感收益仅在同侪影响广度、深度与主动贡献行为间起正向中介效应，而认知收益则在同侪影响深度与反应贡献行为间起负向中介效应。研究拓展了在线网络情境下知识管理与社会学领域的交叉研究，并为在线创新社区社交网络和知识管理提供重要启示。