This study applied the question-oriented problem-solving (QOPS) pedagogy in Statistics learning and addressed several methodological and theoretical gaps in SNA studies to inform instructional and learning implications for teachers and students. The limitations in SNA studies included the lack of theoretical support in the analysis, the homogeneous medium where the data for SNA is collected, the constraints in the diversity of the network modes, and the lack of associations about the SNA measures with learning outcomes. Thus, we investigated 1) how a learning community based on social media evolved with the implementation of the QOPS activities, 2) changes in instructors and learners' roles in light of their one-mode and two-mode SNA measures, and 3) how different relational ties and SNA measures were associated with student outcomes. Participants were 23 graduate students (females: 52.17%) and an instructional team of three. Seven hundred four messages, including 118 posts and 586 comments, were collected from the private advanced statics Facebook learning group over an 18-week advanced Statics course. Aligned with the QOPS approach, the learning community became more interactive and cohesive, as reflected in their one-mode SNA measures. Besides, the two-mode measures, including density, GCC, reciprocity, transitivity, and strong/weak components, demonstrated that the learning community became denser and more interactive via their shared discussion. The instructional team played unique roles as a leader or facilitator in their interaction with students, while students' roles as a leader and coordinator were identified. Students' outdegree, degree, and closeness centralities have the most consistent predictive validity of their learning outcomes. Higher values in degree or closeness via shared discussion topics such as discussion of quizzes, help-seeking, and course administrative communication also predicted performance. The study results have practical implications for instructors and researchers in designing the hybrid learning environment and providing scaffolds to promote help-seeking.

这项研究在统计学学习中应用了面向问题的问题解决方法（QOPS），并解决了SNA研究中的一些方法和理论上的空白，从而为教师和学生提供了教学和学习方面的启示。SNA研究的局限性包括分析中缺乏理论支持，收集SNA数据的同质介质，网络模式多样性的限制以及SNA度量与学习成果之间的关联性不足。因此，我们调查了1）基于社交媒体的学习社区是如何随着QOPS活动的实施而发展的； 2）鉴于教师和学习者的一种模式和两种模式的SNA措施，他们的角色发生了变化；以及3）如何不同的关系和SNA措施与学生的学习成绩相关。参加者为23名研究生（女：52.17％）和3名指导老师。在一个为期18周的高级静力学课程中，从私人高级静力学Facebook学习小组收集了74条消息，包括118个帖子和586条评论。与QOPS方法保持一致，学习社区变得更具互动性和凝聚力，这体现在他们的单模式SNA措施中。此外，包括密度，GCC，互惠性，可及性和强弱组件在内的两种模式的度量表明，学习社区通过共享讨论变得更加密集和互动。教学团队在与学生互动中扮演领导者或促进者的独特角色，同时确定了学生作为领导者和协调者的角色。学生的学历，学位，和亲密关系中心对他们的学习成果具有最一致的预测效度。通过共享讨论主题（例如，测验，求助和课程行政沟通）的程度或亲密程度的更高值，也可以预测表现。研究结果对指导教师和研究人员设计混合学习环境并提供支架以促进寻求帮助具有实际意义。