本科教育作为大学的根和本，其首要职责便是人才培养。近年来，为了解我国本科教育人才培养状况、提升人才培养质量，围绕大学生学习与发展进行的学情调查成为有效的测量和评价手段。其中，学习投入作为大学生学习研究中的“后起之秀”，是近年来调查主持者和教育研究者普遍关注的研究领域。但是，作为一个学术概念，学习投入的内涵和结构还存在诸多模糊之处，需要进行更加细致的审视和更为严谨的检验。 基于此，本研究首先借助绘制国际文献中学习投入研究的知识图谱，描述了学习投入研究的发展趋势、现状特征、研究前沿和知识基础，辨析了高等教育学和教育心理学两方面的研究取向。同时，梳理了其他大学生学习理论在概念建构、结构测量等方面的经验，为学习投入研究的推进提供参考。进而，总结了学习投入的两组内涵维度——“基于个体的学习投入+基于人际的学习投入”及“行为投入+认知投入+情感投入”。在对两组维度进行交叉融合的基础上，提出理解中国大学生学习投入的新框架。 其次，本研究基于“中国大学生学习与发展追踪研究”（China College Student Survey，CCSS）2016-2018年的调查数据，探索并验证了二维六因子的大学生学习投入结构。六个一阶因子分别是：学生个体学习、深层学习策略、积极学习情绪、生师互动行为、反馈调节策略和人际关系感受，它们分别代表大学生基于个体的行为投入、基于个体的认知投入、基于个体的情感投入、基于人际的行为投入、基于人际的认知投入和基于人际的情感投入。在六个一阶因子之上存在“基于个体+基于人际”的高阶结构，但是不存在“行为+认知+情感”的高阶结构。在此基础上，对中国大学生的学习投入状况进行整体描述与群体差异分析。 最后，根据本研究提出的六个学习投入一阶因子，使用聚类分析和判别分析的方法，将中国大学生划分为“高个体高人际型”、“低个体低人际型”、“个体主导型”和“人际主导型”四种学习投入类型。四类学生在人口学变量和高等教育变量上的分布不具有实际差异，主要区别在于对院校支持的感知程度和学生汇报的教育收获。进一步研究发现，在不同类型的学生中，院校支持、学习投入和教育收获之间的影响机制存在差别。

As the root and the foundation of a university, the primary responsibility of undergraduate education is talent cultivation. In recent years, in order to understand the situation and improve the quality of talent cultivation in undergraduate education in China, college student survey has become an important means of assessment and evaluation. During this period, student engagement has turned into a research field of concern to the survey hosts and educational researchers. However, as a theoretical concept, there are still ambiguities in the connotation and conceptual structure of student engagement, which need to be further examined and discussed. This dissertation study, by drawing a knowledge map of international literature on student engagement, describes the history and current development of student engagement research and summarizes the knowledge base of student engagement in the fields of higher education and educational psychology. Furthermore, to provide reference for the further development of student engagement research, this study synthesizes the concept and measurement instruments in other classical theories in college student learning. Based on the literature review, this study proposes a new framework to understand student engagement of Chinese undergraduates. The framework consists of two dimensions: a "individual-based learning engagement + interpersonal relationship-based learning engagement" dimension and a "behavioral engagement + cognitive engagement + emotional engagement" dimension. Secondly, this study explores the structure of student engagement with data from the China College Student Survey (CCSS) and discovers six first-level factors and two second-level factors. The six first-level factors are: individual student learning, deep learning strategy, positive learning emotion, student-faculty interaction behavior, feedback-based regulation strategy and interpersonal relationship, which respectively represent individual-based behavioral engagement, individual-based cognitive engagement, individual-based emotional engagement, interpersonal relationship-based behavioral engagement, interpersonal relationship-based cognitive engagement, and interpersonal relationship-based emotional engagement. At the second level, the “individual-based student engagement” and “interpersonal relationship-based student engagement” dimension is proved for Chinese undergraduate student engagement, but the “behavioral-cognitive-emotional engagement” dimension is not supported by the data. Finally, based on the two-level structure, this study divides Chinese undergraduate students into four types using K-means cluster analysis, namely, “high individual and high interpersonal relationship” type, “low individual and low interpersonal relationship” type, “high individual and low interpersonal relationship” type and “low individual and high interpersonal relationship” type. Discriminant analysis validates this result and derives the equation expression for each type. The study further finds that the influential mechanism between institutional support, student engagement and education outcomes varies across different types of student engagement.