Classification-based development of higher education institutions is an important indicator of the maturation of a higher education system. In the process of the formation and evolution of the classification system for university in China, three structural features have become deeply intertwined: a vertical hierarchy based on institutional levels, a horizontal classification rooted in industrial sectors and academic disciplines, and a resource allocation pattern centered on key construction projects. Their interplay has caused institutional classification to gradually devolve into hierarchical ranking in practice. This predicament is jointly shaped by cultural, fiscal, and governance factors: a monolithic standard of success has eroded institutional type identity; a fiscal logic that privileges selectivity over baseline support has undermined equity across types; and homogenized evaluation systems have rendered classification rules persistently ineffective. To make classification operational, fiscal restructuring should serve as the entry point, establishing category-specific funding guarantees and a differentiated appropriation system while advancing type-based performance evaluation and recalibrating governance instruments, so that institutional reform creates a practical foundation for cultural transition and gradually builds a higher education classification system that accommodates diversity and ensures baseline equity.

Currently, cutting-edge technologies represented by artificial intelligence are advancing rapidly. The innovation-driven development of the AI industry has become a pivotal point in global future industrial competition, urgently requiring a large supply of high-quality talent. Based on data mining of recruitment information from AI enterprises, this paper analyzes the talent demand characteristics of the AI industry from dimensions such as academic qualifications, majors, work experience, professional skills, and interdisciplinary research capabilities. By examining the current state of talent cultivation in AI-related majors at universities, it reveals mismatches in talent supply and demand. Addressing the talent needs of AI industry innovation, the study proposes optimizing AI talent training pathways in higher education, offering insights for universities to refine AI-related disciplines or major setups and improve talent training models.

Focusing on the structural isomorphism risk in the expansion of engineering undergraduate programs and whether interprovincial linkages can foster differentiated specialization, this study uses program filing/approval records from 11 provinces/municipalities in China’s Yangtze River Economic Belt (2015-2024) and applies a Spatial Durbin Model to identify the dynamic evolution and spatial effects of program placement. Results show that within-province industry–education alignment generally increases convergence: once alignment is high, sunk costs and organizational inertia tend to lock institutions into existing trajectories, narrowing the space for structural renewal. Interprovincial spillovers are dominated by competitive imitation, leading to follower-type expansion, and linkage alone does not automatically improve the regional division of labor. By reducing the implementation costs of differentiated allocation, collaborative governance can shift cross-border influences from imitation pressure to specialization incentives, significantly mitigating convergence and high-level lock-in. Accordingly, program governance should move from static matching to frontier-oriented adaptation by strengthening isomorphism early-warning and evaluation mechanisms, promoting interprovincial joint appraisal and differentiated entry for popular fields, and improving dynamic exit and stock-optimization mechanisms.

Drawing on spatio-temporal sociology and using a qualitative longitudinal diary-interview design, this study follows 44 first-year undergraduates from two elite universities over two years to examine the spatio-temporal tensions and divergent trajectories in top-notch students’ high school-to-university transition. The findings show that the transition involves the reconfiguration of cognitive, interactional, and choice spaces, together with a misalignment between institutional and individual time, which places students in a liminal state. Students cross this liminality by reconstructing their individual spatio-temporal order, and their capacity to do so depends on the activation of subjectivity. Three trajectories are identified accordingly: through positive interaction with the institution, students achieve coordination between subjective and objective spatio-temporal orders, cross liminality, and develop subjectivity; when rules are difficult to decode and support is insufficient, students struggle to renew their meaning frameworks, leading to a rupture between subjective and objective spatio-temporal orders, prolonged liminality, and suspended subjectivity development; under performance competition, students carry over their high school inertia, resulting in a mismatch between subjective and objective spatio-temporal orders, an “adaptation trap,” and unactivated subjectivity formation. The study thus calls for process-oriented support systems, a more enabling evaluation ecology, and stronger efforts to activate students’ subjectivity.

The popularization of higher education has heightened academic anxiety among college students. Through three progressive experiments grounded in teaching contexts, this study examined the relationship between academic uncertainty, sense of control, and anxiety. Experiment 1 employed a single-group cross-sectional design to verify the mediating effect. Experiment 2 utilized an experimental group-crontrol group design to manipulate academic uncertainty to confirm its causal impact on anxiety. Experiment 3 adopted an experimental group-control group design to manipulate academic sense of control to test its independent effect on anxiety. Results showed that academic sense of control mediated the relationship between academic uncertainty and anxiety; high academic uncertainty reduced sense of control and increased anxiety; enhancing sense of control alleviated anxiety. These findings reveal a complete influence path: academic uncertainty → academic sense of control → academic anxiety. Accordingly, universities should synergize efforts at two levels: at the source level, promote teaching transparency to clarify objectives, refine assessment standards, and visualize development paths, thereby reducing academic uncertainty; at the mediating level, leverage tool empowerment to support students’ goal decomposition, process regulation, and outcome anticipation, thereby enhancing academic sense of control and systematically alleviating academic anxiety.

In the cultivation of top-notch innovative talents, universities have generally introduced competition systems to “promote innovation through competitions.” However, there is still a lack of micro-level causal evidence on whether and how competition enhances creativity. This study investigated the impact of different competition systems on students’ creativity through creativity assessment based on large language models. The results showed that both indirect and direct competition can significantly enhance students’ creativity, with indirect competition having a more pronounced effect. Further analysis indicated that individual trait anxiety and task difficulty are important variables that affect the role of competition. Among individuals with low trait anxiety, the positive impact of competition on creativity is more prominent; in simple tasks, direct competition is more effective in stimulating creativity, while in complex tasks, the advantage of indirect competition is more evident.

The “loose coupling” phenomenon in traditional industry–education integration has constrained the effectiveness of practice-based education for engineering graduate students. As a practice platform doubly embedded in both academia and industry, supervisor-led academic spin-off provide a new observational setting for examining the mechanisms through which practice-based education is realized. Based on 653 questionnaires, this study analyzes the educational effectiveness of supervisor-led academic spin-off as internship platforms for engineering graduate students. The findings show that, compared with internships in ordinary enterprises, supervisor-led academic spin-off are significantly associated with higher levels of knowledge and skills and with some forms of research output, indicating a certain platform advantage. Distinct participation-role differentiation also exists within the platform, which can be classified into three types: peripheral participants, technically oriented participants, and comprehensive participants; these different roles are associated with differentiated educational outcomes. In addition, educational effectiveness is further shaped by factors such as opportunity allocation, output rules, participation boundaries. So, improving the quality of practice-based training for engineering graduate students should go beyond platform construction and place greater emphasis on platform governance, particularly in task allocation, supervisory mechanisms, and participation pathways.

Drawing on survey data from the National Doctoral Graduate Survey, this study empirically examines the dominant factors and underlying mechanisms of university–industry joint training for engineering doctoral students from the perspective of work-integrated learning. The findings indicate that structural factors—such as the duration of joint training, project type, and firm size—do not exert significant effects on doctoral students’ research and practical competencies. In contrast, factors including the nature of enterprise R&D activities, dissertation topic selection, supervisors’ industry background, and their degree of engagement with industry emerge as key determinants of training outcomes. Further analysis reveals that the effectiveness of joint training is not determined by its duration or organizational form, but rather by the extent to which doctoral research is embedded in real-world engineering R&D contexts. Accordingly, this study suggests that, grounded in the intrinsic logic of work-integrated learning, greater emphasis should be placed on deepening the integration of research tasks within authentic engineering contexts and enhancing the alignment between project characteristics and training processes, thereby facilitating a shift from formal participation to substantive integration. The findings provide empirical evidence and theoretical insights for constructing a work-integrated, university-industry collaborative training mechanism for engineering doctoral education.

The Nobel Prizes in the natural sciences, among the most prestigious international scientific awards, are granted to leading scientists in the field. Examining collaborations among co-laureates from 2014 to 2024 reveals four types of collaboration that can be grouped into two modes: small science collaboration, characterized by small teams and informal intra-institutional cooperation, and big science collaboration, characterized by large-scale, cross-national, and cross-institutional formal cooperation. Drawing on social capital theory and social network analysis, this study compares structural, relational, and cognitive capital across the two modes. The findings show that Nobel Prize–level achievements do not depend on a single optimal collaboration mode but emerge from multiple innovation pathways enabled by different configurations of social capital. Therefore, university research systems should maintain institutional flexibility to support diverse collaboration modes and promote the coordinated development of both small and big science collaboration.

To support the development of the modern industrial system, vocational undergraduate education must clarify the unique logic and practical pathways of talent cultivation based on its type-specific orientation. A theoretical framework for vocational undergraduate talent cultivation is constructed, structured as a “goal orientation-core implementation-empowerment assurance” model across three levels and four dimensions. At the goal orientation level, the focus should be on cultivating technologists, guided by the value orientations of vocational relevance, theoretical depth, innovativeness, and collaboration. Through the systematic reconstruction of key teaching elements-such as programs, curricula, teaching materials, teaching staff, and practical training-the core implementation level is established. Meanwhile, the empowerment assurance level is shaped by the coordinated reconfiguration of supporting elements, including teaching organization, industry-education integration, applied research, governance systems, and quality assurance. Drawing on reform practices at Shenzhen Polytechnic University, the study engages in a two-way dialogue between theory and practice, offering insights for the high-quality cultivation of vocational undergraduate talents.

In the context of higher education classification reform, although the independent type of skill-oriented universities has been confirmed by policy, there is still a theoretical absence regarding the laws of their typological development. From the perspective of historical institutionalism, the “type emergence” of skill-oriented universities has experienced path dependence in the period of institutional inertia, gradual transformation in the period of institutional adjustment, and critical juncture rupture in the period of institutional breakthrough, presenting an institutionalization process from hierarchical attachment to typological independence. Its “typological legitimacy” is historically constructed through long-term interaction with national strategy, social cognition and industrial demand, forming a three-dimensional foundation of policy legitimacy, social legitimacy and functional legitimacy. The inherent stipulation of its “typological development” contains dual logic: endogenous and exogenous. The former is embodied in the logic of institutional operation, knowledge production and skill formation; the latter is embodied in the logic of market response and value realization. To promote the high-quality development of skill-oriented universities, it is necessary to optimize the institutional system from five dimensions: consolidating the institutional foundation, breaking institutional barriers, enhancing institutional embeddedness, driving institutional innovation, and expanding institutional boundaries.