Abstract:

Revealing the intrinsic mechanisms of connectivist learning is a crucial focus for supporting the construction of a new curriculum model centered on connectivist learning, in response to the new propositions of cultivating top-notch innovative talents in the current era. The mining and representation of connectivist learning paths is the essential and cutting-edge research direction within this context. Current research lacks continuous and complete tracking and structural representation of connectivist learning paths, which is hindering the depth of curriculum practice. To solve this problem, based on the first domestic connectivist online course, and focusing on 56 connectivist learners, empirical research was conducted using methods such as cluster analysis, sequence analysis, and network analysis. The study found that, first, learners produced three types of learning paths, namely, main route—problem-oriented, learning support route—community-oriented, and novice route—path-dependent; learners with different experiences have differentiated path preferences, and linear course operation rules affect the space for topic selection. Second, the topological structure of connectivist learning paths has dual characteristics of temporality and networking. Third, learners exhibited three types of learning modes, namely problem-driven, curated connectivity, and knowledge restructuring. Future construction of connectivist online courses and related online learning practices should emphasize the hierarchical classification design and personalized provision of learning resources; design activity chains that stimulate collaborative knowledge creation; enhance dynamic operation and management of online learning practices; and optimize course organizational models, ultimately systematically improving course development strategies across resources, activities, and management dimensions.

Key words: connectivism, cMOOC, learning path