Drawing on the theoretical lens of intertextuality and employing the Latent Dirichlet Allocation (LDA) topic model, this study systematically examines the similarities and differences between Chinese and American science education policies across three comparative dimensions: (1) structural differences in policy design; (2) similarities and divergences in thematic emphasis; and (3) evolutionary trends in policy agendas. The selection of policy texts followed clear criteria: (1) documents must be issued at the national level (central government in China and federal government in the United States); (2) they must be directly related to science education; and (3) local and informal texts were excluded. A total of 23 policy documents were retained, amounting to approximately 200,000 words. The optimal number of topics was set at four for each country, with coherence scores of 0.75 (China) and 3.70 (United States), ensuring model validity. The results indicate a clear divergence in governance logic and semantic structure between the two countries. In terms of topic structure, Chinese policies exhibit a “one-core, multiple-supports” configuration, with state authority at the semantic center and various social actors functioning as collaborative agents, whereas American policies present a “multi-actor co-construction” pattern in which diverse policy texts resonate horizontally. In terms of topic emphasis, both countries display sustained reinforcement of teacher development, forming a shared policy consensus. In terms of topic evolution, China follows a pattern of linear inheritance, continuously assimilating and layering upon existing policy frameworks, while the United States demonstrates semantic migration, with policy meanings shifting through the reconfiguration of key concepts and ideas. As two representative models of education governance, how China and the United States navigate strategic absorption and reciprocal borrowing within their distinct historical, cultural, and institutional contexts holds significant implications for advancing global science education policy from a phase of parallel development to one of mutual innovation.
Li et al. (Thu,) studied this question.