作为影响学业成就的重要非智力因素之一，学习动机一直是教育心理学的研究焦点。学习动机对学业成就的影响途径之一是通过学习投入的中介作用，因此探究学习动机对学习投入的影响机制具有重要意义。诸多认知心理学框架下的学习动机理论从多个角度提出了不同的动机构念，对理解动机的作用机制都具有重要意义，是动机的重要组成成分，并且这些动机成分间存在复杂的相互作用。目前研究中普遍使用的潜变量模型很难同时对多个构念间的复杂相互作用进行系统性的建模，不利于当前对动机成分与学习投入关系的分析。近年兴起的心理测量网络模型能够构建可视化的网络结构，这些网络模型能有效反映多个变量间的复杂关系。通过构建动机成分与学习投入变量的网络模型，可以对动机成分间的相互作用关系及其对学习投入的影响机制进行分析。
本研究从网络模型的视角，以高中学生为研究对象，应用心理测量网络模型中的无向加权相关网络模型，构建学生学习的动机网络模型和动机-投入网络模型，分析动机成分间的复杂关系，并探究学习动机对学习投入的影响机制。
对动机网络的分析结果发现：（1）动机成分间的相互作用能构成复杂的关系网络。自我效能感通过目标定向与兴趣和任务价值建立较强的间接关联，自主选择感与兴趣和任务价值直接联系。动机网络的核心节点是兴趣。（2）亚群体的动机网络比较发现，不同性别和SES的亚群体的动机网络差异不显著，高低成就学生的动机网络差异显著。对低成就学生而言，自主选择感起到的作用相对较大，核心节点是兴趣；高成就学生动机网络的核心节点是掌握目标。
对动机-投入网络的分析结果发现：（1）动机对学习投入的影响强度和方式不同，自我效能感、兴趣、任务价值和掌握目标影响较大，兴趣和任务价值直接影响学习投入，自我效能感和掌握目标对学习投入既有直接联系，又有间接联系，其中间接联系较强。（2）亚群体的动机-投入网络比较发现，不同性别和SES的亚群体的动机-投入网络差异不显著，高低成就学生的动机-投入网络差异显著。对低成就学生，自主选择感对学习投入的影响相比高成就学生较大，核心节点是兴趣；对高成就学生，任务价值对学习投入的影响相比低成就学生较大，核心节点是掌握目标。
本研究提出了一个新的动机整合理论框架，该框架将基于认知动机理论提出的不同动机构念根据认知对象分为对任务的、对自我的和对自我-任务关系的三类动机成分，丰富了动机理论。研究将网络分析方法应用于动机研究，一定程度上克服了潜变量模型的局限性，为研究复杂心理构念提供了新的视角。研究结果揭示了不同动机成分对学习投入的影响机制，发现了核心节点，为在教学实践中制订干预方案提供理论依据。
 

As one of the important non-intellectual factors influencing academic achievement, learning motivation has always been a research focus in educational psychology. One of the ways in which learning motivation affects academic achievement is through the mediating role of learning engagement. Therefore, it is of great significance to explore the mechanisms through which learning motivation influences learning engagement. Various theories of learning motivation within cognitive psychology frameworks have proposed different motivational constructs from multiple perspectives, which are important components of motivation and have complex interactions among them. The commonly used latent variable models in current research make it difficult to systematically model the complex interactions between multiple constructs, which hinders the analysis of the relationship between motivational components and learning engagement. Recently emerging psychological measurement network models can construct visualized network structures that effectively reflect complex relationships between multiple variables. By constructing a network model of motivational components and learning engagement variables, it is possible to analyze the interaction relationships between motivational components and the mechanisms through which they affect learning engagement.
In this study, from the perspective of network models, high school students are taken as the research subjects. The undirected weighted correlation network model in psychological measurement network models is applied to construct a network model of student learning motivation and a network model of motivation-engagement. The complex relationships between motivational components are analyzed, and the mechanisms through which learning motivation affects learning engagement are explored.
The analysis results of the motivational network reveal that: (1) The interactions between motivational components can form complex relationship networks. Self-efficacy has a strong indirect association with goal orientation, interest, and task value, while autonomous choice is directly connected to interest and task value. The core node in the motivational network is interest. (2) Comparing subgroups of the motivational network, there are no significant differences in the motivational network between different genders and SES (socioeconomic status) subgroups, but significant differences are found in the motivational network of high-achieving students and low-achieving students. For low-achieving students, autonomous choice plays a relatively larger role, and the core node is interest. For high-achieving students, the core node in the motivational network is mastery goals.
The analysis results of the motivation-engagement network reveal that: (1) Motivation has different strengths and ways of influencing learning engagement. Self-efficacy, interest, task value, and mastery goals have a significant impact. Interest and task value directly influence learning engagement, while self-efficacy and mastery goals have both direct and indirect effects, with a stronger indirect effect. (2) Comparing subgroups of the motivation-engagement network, there are no significant differences between different genders and SES subgroups, but significant differences are found in the motivation-engagement network of high-achieving students and low-achieving students. For low-achieving students, the influence of autonomous choice on learning engagement is relatively larger, and the core node is interest. For high-achieving students, the influence of task value on learning engagement is relatively larger, and the core node is mastery goals.
This study proposes a new framework for integrating motivation, which categorizes different motivational constructs based on cognitive objects into three categories: task-related motivation, self-related motivation, and self-task relationship-related motivation. This framework enriches motivational theory based on cognitive motivation theory. The study applies network analysis methods to motivation research, which to some extent overcomes the limitations of latent variable models and provides a new perspective for studying complex psychological constructs. The research findings reveal the mechanisms through which different motivational components influence learning engagement and identify core nodes, providing theoretical basis for developing intervention strategies in educational practices.