The two main trends in the development of higher education worldwide are universal access and digital transformation. These trends are bringing about an increase in class sizes and the growth of online higher education. Previous studies indicated that both the large-class setting and online delivery threaten the quality, and the exploration of strategies to ensure quality teaching and learning in the large-class setting was in face-to-face or blended learning mode. This study contributes to this topic by exploring the quality of teaching and learning in a new scenario: the fully online large university class. Furthermore, it proposes to use student engagement as a new means to explore the quality of teaching and learning in a large-class setting as it offers evidence on quality from the in-itinere perspective rather than the more commonly ex-post perspective offered by existing studies, collected, for example, from student feedback or course grades. This study was conducted in a mandatory course at an Italian university. Both the Moodle log data and students’ reflective diaries are collected to analyze the presence of students’ behavioral, emotional, and cognitive engagement. Tableau and NVivo handle the quantitative and qualitative data, respectively. By confirming the presence of all three types of engagement, the result indicates quality teaching and learning happens in the fully online large university class. Since we select both “high-grade” and “low-grade” students as representative samples, the Tableau visualization also indicates that only using behavioral engagement to predict students’ academic performance is unreliable.

全球高等教育发展的两大趋势是普及和数字化转型。这些趋势带来了班级规模的扩大和在线高等教育的增长。以往的研究表明，大班教学和在线授课都对教学质量构成威胁，大班教学中保证教学质量的策略探索主要是面授或混合式学习模式。本研究通过探索新场景（完全在线的大型大学课堂）中的教学质量来为该主题做出贡献。此外，它建议使用学生参与作为探索大班环境中教学质量的新手段，因为它从过程中的角度提供质量证据，而不是现有的更常见的事后角度。研究，例如从学生反馈或课程成绩中收集的。这项研究是在一所意大利大学的必修课程中进行的。收集 Moodle 日志数据和学生的反思日记来分析学生的行为、情感和认知参与情况。Tableau 和 NVivo 分别处理定量和定性数据。通过确认所有三种类型的参与的存在，结果表明在完全在线的大型大学课堂中发生了高质量的教学和学习。由于我们选择了“高年级”和“低年级”学生作为代表性样本，Tableau 可视化还表明，仅使用行为参与度来预测学生的学业成绩是不可靠的。