Sage Journals: Discover world-class research

Abstract

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.

Keywords

science education policies comparative analysis Sino-American

Introduction

The global community is well aware of the ongoing competition between China and the United States (X. Wu, 2020). At its core, this competition can be defined as a struggle for dominance in the field of technology (Zhao, 2019). As a foundation for cultivating future technological talent, strategic choices or formulations in science education policy not only influence the trajectories of innovation capabilities and technological strengths in both countries but also affect their positions within the technological innovation chain, potentially impacting the global development landscape (Bianchi & Giorcelli, 2020; Marginson, 2018; Omenn, 2006). In this regard, this paper conducts an in-depth analysis of the science education policies of China and the United States, elucidating the policy orientations of both countries and the underlying logic of development. In this way, the analysis provides a reference for countries around the world when formulating science education policies.

The discrepancies between the science education policies of China and the United States are not merely a matter of contrasting policy texts; rather, they are reflections of the distinct social, cultural, historical, and economic contexts that shape each country’s educational policies (Chen et al., 2021). This aspect is of paramount importance to the comprehensive examination of Sino-American science education policies in this study, which is based on intertextuality theory. The primary contribution of intertextuality theory is its emphasis on the interaction between textual content and form, examining texts within a broader cultural context and highlighting the relationship between text, culture, and practice (Nöth, 1990, p. 9). In the field of education, scholars both domestically and internationally have begun to explore the potential applications of intertextuality theory in educational research. However, its application remains relatively limited, primarily concentrated in the area of teaching research (Bezemer & Cowan, 2022, pp. 107–118; Knain et al., 2021; Oechsler & Borba, 2020). In light of this, this paper aims to expand the application scope of intertextuality theory to a more macro level of education, thereby providing a novel theoretical perspective for the analysis of Sino-American educational policy texts.

This study has selected China and the United States as sample countries for policy analysis, employing an intertextual perspective in a novel manner to construct a new framework for comparing educational policies between the two nations. The objective of this research is to utilize the LDA model for the textual analysis of science education policies in order to answer two core questions. Firstly, what are the differences in the policy orientations of science education between China and the United States? Secondly, how do these policy orientations manifest similarities and differences over time? Through a comprehensive examination of these inquiries, it was found that the evolution and prospective trajectories of science education policies in both countries facilitate the more precise identification of their positioning in the global technology competition and the formulation of educational strategies that align with their developmental needs in other nations.

Literature Review

Amid intensifying global competition in science and technology, science education has increasingly become a key pillar of national soft power and capacity for innovation. Despite the significant differences between China and the United States in terms of educational systems, cultural traditions, and social structures, the two countries have gradually converged in their overarching policy goals for science education, both identifying the enhancement of public scientific literacy and the strengthening of teacher professional development as core priorities (Yan et al., 2024). This shared consensus has driven both nations to introduce a series of national-level science education policies over the past decade. In China, the launch of a new round of science curriculum reform in 2017 is widely regarded as a watershed moment for systematizing science education at the basic education level (Yao & Guo, 2018). In the United States, the STEM Education Strategic Plan (2018–2023) emphasized the cultivation of real-world problem-solving competencies through interdisciplinary integration and project-based learning (Tan, 2015; Tao et al., 2013). However, despite their similar goals, the two countries exhibit starkly different institutional characteristics in policy implementation mechanisms and reform pathways.

China, operating under a centralized governance system, tends to adopt a “national top-level design—local tiered implementation” model to facilitate rapid policy rollout. While this model is advantageous in terms of efficiency and broad coverage, the complexity and scale of China’s education system, coupled with pronounced regional disparities, often lead to a disjunction between policy ideals and implementation realities (Liang et al., 2017, pp. 2–3, 10). In contrast, the United States relies on a pluralistic governance structure wherein governments, schools, enterprises, and civil society organizations jointly participate in curriculum reform and resource provision (Aikenhead, 1997). This model emphasizes innovation and autonomy, but the absence of a strong central coordinating mechanism creates challenges for policy continuity and coherence (Johnson, 2012). Thus, institutional structures fundamentally shape the logic of policy implementation: China prioritizes uniformity, while the United States emphasizes diversity.

Nevertheless, institutional design is not the sole determinant of policy outcomes. Cultural traditions exert a profound influence on teacher role perception and pedagogical paradigms. Rooted in the Confucian ethos of respecting teachers and valuing education, Chinese teachers are traditionally perceived as knowledge authorities and moral exemplars (Cheng & Wan, 2015; F. Liu, 2023). This role orientation has long entrenched teacher-centered classroom practices. Even when policy reforms promote student-centered approaches, their enactment is constrained by prevailing cognitive frameworks (E. Liu et al., 2015). Notably, Cheng and Wan (2015) found that Chinese students are not passive recipients; rather, they excel in deepening understanding through repetitive learning, albeit within teacher-led knowledge structures. In contrast, American educational culture emphasizes egalitarian interaction and individual agency, positioning teachers as learning facilitators and resource coordinators whose primary task is to stimulate inquiry and critical thinking rather than deliver standard answers (Foley & Reveles, 2014). Such cultural differences result in divergent “adaptation outcomes” when similar pedagogical concepts are adopted. As Tan (2015) notes, China often exhibits surface-level appropriation of Western educational ideas—incorporating terms like “project-based learning” and “collaborative learning” at the discursive level while retaining teacher-dominant structures in classroom practice. Thus, culture not only shapes teaching behaviors but also defines the boundaries of localization in policy borrowing.

The structure of resource allocation is likewise an important variable influencing policy effectiveness. Numerous studies have revealed a significant developmental gap in science education between China’s eastern and central-western regions. High-quality teachers, laboratory facilities, and extracurricular opportunities are disproportionately concentrated in economically developed areas, while rural and underdeveloped regions face notable constraints in participating in scientific activities and curriculum innovation (Cong et al., 2021). Gender disparities also present hidden barriers in policy implementation. Although dedicated funding programs for women in science have been introduced, Ming (2023) argues that traditional gendered role expectations continue to influence women’s decisions to participate in scientific fields. In the United States, despite its overall advantage in educational resources and relatively favorable technological conditions, the highly decentralized governance structure places low-income and minority students at a systemic disadvantage in accessing science education opportunities (Smith et al., 2016). Johnson (2012) further points out that the lack of cohesive funding and coordination mechanisms often results in policy disruption or regression during implementation.

In sum, as two archetypal education governance models, China’s centralized policy mobilization system and the United States’ decentralized policy coordination mechanism diverge structurally in value orientations, institutional logic, and implementation strategies. How the two countries—within their distinct historical, cultural, and institutional contexts—can achieve reciprocal absorption of policy discourses and strategic borrowing, thereby advancing science education policies from parallel development to mutual innovation, has become a pivotal question in international science education policy research. To address this, the present study adopts intertextuality theory as an analytical lens, aiming to uncover the internal connections and divergent mechanisms in the thematic structures, discursive expressions, and semantic transfers of Chinese and American science education policy texts, thus providing more explanatory evidence for cross-national policy learning.

Research Design

Theoretical Foundation

Kristeva, a French literary critic and theorist, introduced intertextuality theory in 1966. Its core idea is that “intertextuality emphasizes that any work is essentially a ‘mosaic’ composed of the absorption and transformation of many other texts” (Allen, 2011, p. 39). Texts are connected through means such as citation, implicit reference, indirect mention, and imitative creation, allowing one text to reflect certain characteristics of another. Intertextuality theory posits that a text contains both an internal and an external structure. At the level of internal structure, Kristeva suggests that “one word is the reappearance of another” (as quoted in Samoyault, 2003, p. 136). A word within a text is not merely an independent symbol; it carries its own semantic connotations, conventional usages, and normative functions. However, when placed in combination with other words, its meaning, usage, and normative orientation often shift in order to adapt to the overall context of the text. At the level of external structure, Culler argues that intertextuality refers to a process through which one text absorbs, imitates, or critiques other texts—constituting a relationship of intellectual inheritance and development (Worton & Still, 1990, p. 98). Such a relationship is not fixed or static; rather, it is open, fluid, and indeterminate, forming a dynamic network system. As texts continuously imitate, adapt, and reference one another, they become interwoven into an ever-expanding network. Kristeva further contends that textual analysis must be situated within a broader space—“to interpret the text within social and historical (texts)” and “to read it across both synchronic and diachronic dimensions”—so as to stimulate the continuous emergence of new meanings (as quoted in Raj, 2015).

By recognizing texts as independent research objects, intertextuality theory opens up a new field of study for academics. By introducing this idea into text analysis, comparative methods have been developed, which provide a deeper analysis of linguistic features and academic thought through comparing similarities and differences among texts regarding language, viewpoints, and ideas (Yang, 2023). As a matter of fact, policy texts are not isolated from the rest of society. They are embedded within an intertextual network through absorption, imitation, borrowing, and critique, and are influenced by historical, cultural, and political contexts. Current policies are often built on the inheritance and development of past policies, and this continuity and evolution make the interaction between policy texts particularly critical. Based on intertextuality theory, Sha (2020) explored how elite school resources promote regional educational equity through an analysis of historical texts on elite schools. Therefore, applying intertextuality theory to policy text analysis is feasible.

Building on this foundation, the LDA (Latent Dirichlet Allocation) topic model provides a technical pathway for operationalizing intertextuality theory. Intertextuality theory posits that within policy texts, the basic symbolic unit is not the word but the theme. LDA, which centers on word co-occurrence probabilities, is designed precisely to identify the latent semantic structures across texts. It assumes that each document consists of several latent topics, and each topic is represented by a cluster of high-frequency co-occurring words (Chen et al., 2016; Jelodar et al., 2019). In doing so, it transforms the “semantic relation network” emphasized in intertextuality theory into a quantifiable topic distribution matrix. This methodological integration enables the relational philosophy of intertextuality to be quantitatively manifested in computational text analysis, thereby achieving an organic unification of theory and method. Based on LDA topic modeling, this study uncovers the recurrent core issues, policy structures, and semantic divergences across different time periods or national contexts within policy texts, ultimately mapping the “intertextual network” of semantic relations in Chinese and American science education policies.

LDA Model

The LDA model, as a text mining tool, was proposed by Blei et al. (2003). It is a three-layer Bayesian model of “document-topic-word,” which introduces Dirichlet prior distributions based on the vector space model and statistical language models. The LDA model uses probabilistic distributions to analyze the topics within each text and then classifies the corpus according to those topics.

Assuming that the distribution of topics (z) and words (w) follows a multinomial distribution, the joint distribution of LDA topic model variables can be represented as follows:

$\begin{matrix} P (w, z, θ, φ, α, β) = & Π_{j = 1}^{M} {P (θ}_{j}; α) Π_{i = 1}^{K} P (φ_{i}; β) \\ Π_{t = 1}^{N} P (z_{j, t} | θ_{j}) P (w_{j, t} | φ z_{j, t}) \end{matrix}$

In the above equation, α and β are hyperparameters: the former represents the distribution of topics over documents, and the latter represents the distribution of words over topics; φ and θ denote the multinomial distribution that governs the joint distribution of topics and words; Z represents the topic assigned to each word according to θ; and W refers to the sequence of words generated through sampling based on the word distribution φ and Z. By connecting these words, a document is generated. By repeatedly generating a large number of documents in the corpus and comparing them with the original documents, the optimal arrangement of points in the implicit Dirichlet distribution can be identified to maximize the probability.

Data Processing

Sources and Screening of Policy Texts

Policies are political actions or prescribed guidelines taken by governments, political parties, and other groups to achieve specific political, economic, cultural, and social goals during a certain period. This includes strategies, laws, decrees, measures, methods, and regulations (Su, 2014, p. 19). To systematically collect and analyze relevant policy documents on science education in Chinese primary and secondary schools, this study followed the following strategies: First, the authors searched for relevant laws, methods, and regulations related to science education in primary and secondary schools in the government documents category of China National Knowledge Infrastructure (CNKI). Second, the authors searched for relevant policy documents through official government databases such as the State Council, Ministry of Education, and Ministry of Science and Technology. Finally, the authors used artificial intelligence search engines like Baidu, Kimi, and ChatGPT for supplementary searches. This study collected a total of 22 policy documents and further screened and supplemented them according to the following criteria: (1) The issuing units are the central government of China (ministries and directly affiliated agencies of the State Council); (2) The documents are closely related to science education, with content directly stipulating and reflecting the management, content, and measures of science education; (3) This study does not include speeches, work reports, and other policy documents by national leaders and education authorities (Huang et al., 2015). Based on the preliminary collection of policy texts, this study, considering the timeliness, completeness, and authority of the policies, strictly selected and finally determined 15 policies as research samples, as detailed in Table A1 (see Appendix A).

To systematically collect and analyze relevant policy documents on science education in American primary and secondary schools, this study used “Science Education” as the keyword for searching. It was found that there were very few relevant policy texts, mainly concentrated on “The Next Generation Science Standards.” Science education is part of STEM education. Therefore, this study further used “STEM”“STEM Policy” and “STEM Act” as keywords to search databases on official U.S. government websites such as the United States Congress, the U.S. Department of Education, and the National Science and Technology Council. Secondly, open databases such as Kaggle, Paper with Code, and Sci-hub were used to identify relevant policy documents. Finally, supplementary searches were conducted using artificial intelligence search engines such as Baidu, Kimi, and ChatGPT. In this regard, this study collected a total of eight relevant policy documents and, according to the above selection criteria, selected eight policies as research samples, as detailed in Table B1 (see Appendix B).

Text Preprocessing

During the construction of the LDA topic model, text preprocessing is an important step that directly impacts the accuracy of the subsequent data processing. This study adopts the following steps:

Enhance Relevance

As a result of the study, content has been removed from China-U.S. policy texts that are not related to science education.

Natural Language Processing

This study uses the Jieba and NLTK libraries in Python for tokenization, stopword removal, and part-of-speech enhancement to process the science education policy texts from China and the United States, forming the text corpora. In addition, for the English corpus, words shorter than two characters or longer than five characters were excluded.

Feature Enhancement

To construct a term frequency matrix, this study uses the doc2bow technique from the Gensim library. To address the issue that traditional tokenization techniques cannot accurately identify the actual importance of words in the text, the TF-IDF model is introduced to enhance the model’s tokenization discrimination capability (Gan & Qi, 2021).

Manual Review

The authors manually filtered out disruptive content in the text, such as garbled text, special punctuation and symbols, numbers, etc.

LDA Modeling

Before running the LDA model, it is necessary to determine the number of topics. Topic coherence refers to the degree of similarity between words within the same topic in the corpus. The study finds that using coherence scores as a criterion for determining the number of topics is effective, based on existing research results (Y. Wu et al., 2024). Generally, the higher the coherence score, the better the performance of the topic model. During operation, the Dirichlet priors were set to the default values of the toolkit (α = 1/K, β = .01), and the number of training iterations is fixed at 50.

Number of Topics in Chinese Science Education Policies

As shown in Figure 1, Coherence reaches its peak when K = 4; thus, the inflection point of 4 is the optimal value for K. To further validate and illustrate this, the pyLDAvis library was used to visualize the LDA topic model, as shown in Figure 2. When K = 4, the boundaries between topics are clear and the classification effect is better, so 4 is chosen as the number of topics.

Figure 1.

Number of topics—coherence line chart (China).

Figure 2.

PyLDAvis visualization results (K = 4).

The Number of Themes in U.S. Science Education Policy

From Figure 3, it can be seen that when K = 4, the coherence score is the highest, making the inflection point at 4 the optimal value for K. To further validate and explain this, the pyLDAvis library was used to visualize the LDA topic model, as shown in Figure 4. When K = 4, the boundaries between the topics are clearly defined, and the classification effect is optimal, which is why four topics were chosen.

Figure 3.

Topics number-consistency line graph (United States).

Figure 4.

PyLDAvis visualization results (United States).

Research Results

When policy texts are deconstructed into individual lexical units, similar terms are aggregated into specific topics that collectively reflect the underlying intentions of the policy. This process itself constitutes an act of intertextual construction (Johansen & Larsen, 2005, p. 67). This process not only involves interpreting the text internally, but also situating it within broader contexts and related elements, thereby forming an orderly and structured configuration. Therefore, identifying the themes clearly is particularly crucial. To this end, after training the LDA topic model, this study extracts the generated “topic-word” distribution. Based on the high-probability characteristic words under each topic, the content of the topics is interpreted, and topics are identified and labeled.

Thematic Analysis of Science Education in the China

This paper extracts the “topic-word” distribution generated from the Chinese science education policy texts and identifies the topics based on the characteristic words of four themes. These topics are identified as: teaching practice, teacher development, popular science work and cultivation of student ability in science education, as shown in Table 1.

Table 1.

Topics-Keyword Distribution of Policy Texts (China)

Number	Thematic identification categories	詞項(主題前10個高概率特徵詞)	The term (top 10 high-probability characteristic words for the topic)
Topic 1	Teaching practice in science education	科普、加強、教育、科學、科技、開展、中小學、教學、學生、工作	Popular science, strengthen, education, science, science and technology, carry out, primary and secondary schools, teaching, students, work
Topic 2	Teacher development in science education	教育、科學、實驗教學、教師、課程、活動、學生、科普、實驗、加強	Education, science, experimental teaching, teachers, curriculum, activities, students, popular science, experiments, strengthen
Topic 3	Science popularization in science education	科普、科學、科技、工作、建設、發展、創新、素質、社會、開展	Popular science, science, science and technology, work, construction, development, innovation, quality, society, carry out
Topic 4	Cultivation of student ability in science education	學生、科學、學習、課程、評價、教學、核心、探究、實踐、教師	Students, science, learning, curriculum, evaluation, teaching, core, inquiry, practice, teachers

Thematic Analysis of Science Education in the United States

This paper extracts the “topic-word” distribution generated from the U.S. science education policy texts and identifies the topics based on the characteristic words of four themes. These topics are identified as: government and social support, teacher development, teaching practice and cultivation of student ability in science education, as shown in Table 2.

Table 2.

Topics-Keyword Distribution of Policy Texts (United States).

Number	Thematic identification categories	The term (top 10 high-probability characteristic words for the topic)
Topic 1	Government and social support in science education	STEM, education, federal, program, community, agency, plan, support, practice, science
Topic 2	Teacher Development in Science Education	science, student, program, education, teacher, mathematic, STEM, knowledge, base, school
Topic 3	Teaching practice in science education	STEM, science, education, student, knowledge, technology, practice, program, scientific, learning
Topic 4	Cultivation of student ability in science education	STEM, student, learn, learning, school, technology, access, education, provide, tool

Based on the above findings, China and the United States have commonly developed a tripartite consensus of “instruction–teachers–students” within their science education policies. In contrast, “science popularization” occupies a prominent position in China’s semantic network, whereas “government and societal support” emerges as the central semantic focus in the United States.

Comparison of Science Education Policy Orientations Between China and the United States

Once the topics are clarified, this paper uses them as basic symbols to explore three aspects regarding the differences in science education policy orientations between China and the United States: topic structure, focus, and evolution.

Topic Structure

As symbolic units within policy texts, topics constitute a distinctive semiotic system, in which the topic structure refers to the overall configuration and interrelationships among all topics within a policy document. The meaning of a topic structure is determined not only by the topics themselves, but also by their relationships with other policy topics and their position within the broader policy system. As such, this paper uses Python tools to visualize the topic structure, providing an intuitive presentation of the thematic structure of Chinese and American science education policy texts, as shown in Figures 5 and 6.

Figure 5.

Topics structure diagram (China).

Figure 6.

Topics structure diagram (United States).

Node Differences

In China's topic structure diagram, the nodes mainly focus on “innovation,”“promotion,” and “science popularization,” indicating that China’s policy is guided by an emphasis on technological innovation. At the same time, nodes such as “education,”“experiment,” and “students” show that China’s policy also places a high emphasis on teaching. In the U.S. policy structure diagram, the nodes mainly focus on “practice,”“technology,” and “learning” demonstrating U.S. policy’s emphasis on practice in science education. Additionally, nodes such as goals, plans, and community indicate that U.S. policy also promotes diverse stakeholders.

Connectivity Differences

In China’s topic structure diagram, the connectivity between nodes is strong, particularly the connection between “innovation” and “promotion” reflecting the high relevance of “innovation” and highlighting its central role. In the U.S. topic structure diagram, the connectivity between nodes is also strong, especially between “practice” and “technology” as well as “learning” and “education,” indicating that U.S. policy focuses on practice, technology application, and learning methods in science education.

Based on the above analysis, although both countries maintain a high focus on teaching practice, the semantic chain of “innovation—promotion—education” in Chinese policy texts reflects an education philosophy driven by innovation. In contrast, the semantic cluster of “practice—technology—learning” in U.S. policies embodies an educational logic centered on practical experience.

Topic Focus

Topic emphasis is assessed by evaluating the probability of word occurrences within each topic, which helps identify the primary concerns and priorities of a policy. This facilitates a clearer understanding of the major issues and focal points in Chinese and American science education policies. Accordingly, this study reveals the relative importance of each topic within the texts by examining the intensity of topics generated by the LDA model. Higher topic strength indicates a greater proportion of that topic, as detailed below:

The distribution of topic strength in Chinese science education policy texts is as follows: Topic 1 is 0.122052, Topic 2 is 0.459134, Topic 3 is 0.340007, and Topic 4 is 0.078807. Among them, Topic 2, which focuses on teacher development in science education, stands out the most.

The distribution of topic strength in U.S. science education policy texts is as follows: Topic 1 is 0.218319, Topic 2 is 0.273254, Topic 3 is 0.251502, and Topic 4 is 0.256926. The strengths of the four topics are relatively close. Among them, Topic 2, which focuses on teacher development in science education, is more prominent.

To further explore the reasons behind the topics, this paper conducts a frequency analysis of the prominent topics in science education policy texts of China and the U.S. and lists the top 10 keywords by weight (Tables 3 and 4).

Table 3.

Word Frequency Statistics (China).

Number	Word	Frequency	Weight	Proportion (%)
1	教育(Education)	533	0.335190	3.52
2	科學(Science)	470	0.172965	3.10
3	實驗教學(Experimental teaching)	233	0.087296	1.54
4	教師(Teachers)	209	0.079475	1.38
5	課程(Curriculum)	193	0.074270	1.27
6	活動(Activities)	152	0.035854	1.00
7	學生(Students)	147	0.040763	0.97
8	科普(Popularization of Science)	141	0.045702	0.93
9	實驗(Experiments)	135	0.041972	0.89
10	加強(Strengthen)	130	0.049241	0.86

Table 4.

Word Frequency Statistics (United States).

Number	Word	Frequency	Weight	Proportion (%)
1	Science	625	0.456875	3.82
2	Teacher	356	0.260236	2.18
3	Program	237	0.173247	1.45
4	Education	234	0.171054	1.43
5	Student	181	0.132311	1.11
6	Federal	155	0.113305	0.95
7	STEM	144	0.105264	0.88
8	Knowledge	135	0.098685	0.83
9	Base	135	0.098685	0.83
10	School	114	0.083334	0.70

As can be seen in Tables 3 and 4, both China and the United States have “teacher development” as a core semantic node in their science education policies. Whether it is the “teacher—curriculum—experimental teaching” in Chinese policies or “teacher—program—education” in American policies, both highlight the central role of teachers in the science education system.

Topic Evolution

Topic evolution can be understood as a form of relational “transfer,” whereby, through time-window analysis, one can observe changes in the activity level and salience of certain topics across different periods (Scholes, 1982, p. 102)—for instance, the emergence and decline of topics, as well as the migration or diffusion of thematic content into other topics. To this end, this study uses Python to plot line charts of topic strength evolution, visually presenting the trends of strength changes for different topics over various time stages, revealing the dynamic changes and development trends of science education policies in China and the United States, thereby providing robust evidence and insights for the research, as shown in Figures 7 and 8.

Figure 7.

Topics intensity evolution chart (China).

Figure 8.

Topics intensity evolution chart (United States).

From the perspective of these themes, both China and the United States show an upward trend in Theme 1 and a downward trend in Themes 2, 3, and 4. The details are as follows:

Theme 1 shows an upward trend in both China and the United States, indicating that this theme has been reinforced or refocused in the new policy cycle. For instance, China’s policies have placed increasing emphasis on teaching practices in science education, while the United States has focused on government and societal support for science education.

Themes 2, 3, and 4 show a downward trend in both countries, indicating that these themes have become relatively stable at the institutional level, the policy goals have been partially achieved, or they have been absorbed by higher-level policy themes (such as innovation-driven development, educational equity, or technological advancement). For example, China’s policies have seen a decline in attention to popular science education, student cultivation, and teacher development, while the U.S. policies have decreased their focus on student cultivation, teacher development, and teaching practices.

Discussion

The purpose of this study is to reveal the similarities and differences in science education policies between China and the United States, providing valuable references that other countries can use develop effective science education policies. Grounded in intertextuality theory, this study employs the LDA topic model to analyze 23 science education policy texts from both countries across three dimensions—“topic structure,”“topic focus,” and “topic evolution.” The study aims to answer the following two core questions: First, what are the differences in the policy orientations of science education between China and the United States? Second, how do these policy orientations manifest similarities and differences over time?

Intertextuality of Theme Structure

The meaning of a symbol does not lie in the symbol itself but in its relationality; it depends on the differences between symbols and their position within the overall semiotic system (Johansen & Larsen, 2005, p. 67). As Saussure emphasized, a symbolic system is composed of interdependent and mutually conditioned meanings, forming a unique internal order. Although both China and the United States emphasize instructional practice in their science education policies, the internal semiotic order they construct differs fundamentally. In China, instructional practice is regarded as a key pathway for cultivating innovation, which is closely associated with the “core–periphery” structure of Chinese science education policy discourse. Party leadership constitutes the central semantic core of the policy, explicitly asserting that “innovation is the primary driving force of national development,” while the collaboration among families, schools, and society occupies a “peripheral support” position within the semantic network. This structure is not merely a technical linkage between policies but rather a semantic “redistribution”: Party leadership provides a unified value framework and directional discourse, whereas the triadic collaboration of home–school–society endows the policy with openness and practicality, extending science education from a “national strategic mission” to a “societal co-education endeavor” (Li, 2025).

In contrast, instructional practice in the United States is conceptualized as a process of cultivating real-world problem-solving capabilities, corresponding to the “multi-centered co-construction” structure of the American policy discourse. Within a policy system where federal agencies, state governments, and social organizations operate in parallel, semantic connections among policies are not unidirectional transmissions but rather horizontal resonances. The diverse textual sources issued by federal, state, and civic actors respond to and reference one another within the policy network, transforming “science education” from an internal school-based disciplinary practice into a form of “public action.” As sociologist Pickering argues, “science-in-practice”—understanding the nature of science through social engagement—has fostered a paradigm shift in American science education, moving from science as a technical practice to science as a civic practice (Musschenga & Gosling, 2005). This tendency is particularly evident in the Framework for K–12 Science Education, which signals a transition from “professional science” toward “everyday practice” (Shamos, 1995). Such a networked intertextual structure highlights the openness and collaborative nature of U.S. science education policies, allowing them to be continuously “rewritten” through societal practice.

Intertextuality of Theme Focus

Kristeva asserted that “a text is a productive force,” suggesting that a text engages in a destructive-constructive redistributional relationship with its environment (as quoted in Samoyault, 2003). From this perspective, the continuous reinforcement of the “teacher development” theme in Chinese and American science education policies represents precisely such a process of deconstruction and reconstruction, through which the policy texts acquire new meanings and redistribute existing ones, thereby realizing productivity. Research has shown that high-quality teachers can enhance students’ self-efficacy and achievement (Sansone, 2017). However, the shortage of science teachers in both countries has constrained the development of science education. In China, a large-scale survey across 31 provinces revealed a severely imbalanced structure within the elementary science teaching workforce. The majority of science teachers serve in part-time roles, accounting for 70.1%, while only 27.5% have a science-related academic background; those with graduate degrees represent merely 1.8%, and novice teachers with less than 5 years of experience constitute as high as 60.6% (Qiao, 2025). In the United States, 85.4% of teacher shortage reports across states in 2025 indicated shortages in science teachers (Learning Policy Institute, 2025). Recent studies have shown that teacher shortages in STEM fields are particularly severe and continue to worsen (García & Weiss, 2019; Nguyen et al., 2022). Furthermore, schools and districts report that filling STEM positions is three to four times more difficult than filling other teaching roles (Goldhaber et al., 2022).

Such challenges often necessitate the reallocation of existing resources and the re-prioritization of policy agendas, reflecting policymakers’ interpretations of social problems and developmental needs, and ultimately demonstrating the productivity of policy texts. The intertextual relationship surrounding the “teacher development” theme in Chinese and American policy documents reveals how science education policies, through teachers as “mediating texts,” enable the continuous reproduction of educational innovation. This convergence further reflects the shared trajectory of global science education policy discourse.

Intertextuality of Theme Evolution

Kristeva argues that in analyzing the relationship between a text and its pre-text, one must account for the historicity embedded in the process of textual formation (Samoyault, 2003). In other words, Chinese and American science education policies should be understood as improvements, developments, and evolutions of their preceding policies—each policy adjustment constitutes a reinterpretation and semantic displacement of earlier policy texts. However, the two countries diverge fundamentally in their paths of semantic transfer.

In China, the theme of “teaching practice” has been continuously reinforced, while themes such as science popularization, student development, and teacher development have gradually weakened. This reflects a “linear inheritance” transfer path in Chinese science education policy, whereby new policies absorb the linguistic framework of previous ones (e.g., “innovation-driven development,”“scientific literacy enhancement”), thereby retaining semantic continuity. Yet as policy enters an institutionalized phase, certain themes (such as science popularization or science teachers) are gradually “embedded” into higher-level strategic texts (e.g., technological innovation strategies or the “Education Power” initiative). Although their explicit visibility declines, they undergo a substantive semantic shift—from “independent themes” to “internalized elements.” Since the release of the Outline of the National Action Plan for Scientific Literacy (2021–2035), the state has expressed clear support for science popularization, yet has not issued a dedicated science education policy at the operational level (S. Wang et al., 2022). This has to some extent constrained the deepening of science education, suggesting that future national policies will likely move toward the formulation of specialized science education policies. The Opinions on Strengthening Science Education in Primary and Secondary Schools in the New Era (2023) explicitly identified science experiment teaching as a key direction of classroom reform, leading to a rise in attention to the “teaching practice” theme. In comparison, although the themes of science popularization, student development, and teacher development appear to be declining, this is not due to policy neglect but rather to the maturity of existing institutional systems, whereby the policy focus has shifted from “construction” to “embedded operation.”

In contrast, U.S. policies exhibit an upward trend in the theme of “government and societal support,” while themes such as student development, teacher development, and teaching practice gradually weaken. This reflects a “semantic migration” transfer path, wherein policy texts across different stages achieve meaning crossover, diffusion, and elevation through shared keywords and cross-referenced concepts. The STEM Education Strategic Plan (2018–2023) focused on curriculum design, teacher training, and resource allocation. Once these tasks were completed, the U.S. government issued the STEM Education and Workforce 2030 Vision in 2023, shifting emphasis toward cross-disciplinary issues such as artificial intelligence, climate change, and sustainability. This marked a new phase of science education policy oriented toward innovation ecosystems and societal impact, wherein the semantic focus migrated from “internal educational improvement” to “cross-sectoral integration and innovation.” Such semantic migration underscores the openness of the intertextual policy network and reflects the multi-stakeholder governance structure within U.S. science education, which in turn drove the rise of the “government and societal support” theme. Since 2020, the National Science Foundation (NSF) has strengthened collaboration between government and civil society through initiatives such as the STEM Teacher Leadership Network and Advancing Informal STEM Learning (AISL), promoting the integration of formal and informal science education. Meanwhile, the themes of student development, teacher development, and teaching practice have begun to decline.

Intertextual Differences

The intertextual differences between Chinese and American science education policies reflect the deeper divergences in their respective models of educational governance and policy discourse systems, which fundamentally stem from disparities in policy-making logic, governance structures, and epistemic rationalities. Regarding differences in policy-making logic, in China, science education policies are embedded within the broader framework of national strategic planning. Most policy documents are initiated by central authorities, drafted by educational administrative agencies, and disseminated hierarchically. This institutional logic shapes a vertical semantic structure of “integration–transmission–implementation.” In contrast, policy formation in the United States results from multi-stakeholder negotiation among federal and state governments, research institutions, civil society organizations, businesses, and even local communities, forming a horizontally co-constructed system (Deng, 2024; Salvatierra & Cabello, 2022). Where differences in governance structure are concerned, Chinese policies display a pronounced dominance of state discourse. Science education is embedded within national modernization and innovation-driven agendas, and intertextuality among policy documents primarily serves political objectives and strategic extension. As a result, semantic shifts tend to occur within the internal policy system (e.g., from “science popularization” to “teacher development”), representing intra-systemic reproduction. In the United States, however, intertextuality in policy discourse reflects the competitiveness of societal discourses. Semantic updates in education policy are more often driven by the redefinition of social issues (such as equity, participation, or community responsibility), meaning that intertextuality functions as a process of social semantic negotiation (Staus et al., 2021; Zhu, 2011, p. 18). Regarding differences in epistemic rationality, the semantic logic of Chinese science education policy aligns more closely with normative rationality, being oriented toward the realization of national goals and emphasizing consistency in systems, standards, and directives. The evolution of policy semantics thus represents an extension from macro-rationality to institutional rationality. In contrast, American science education policy embodies practical rationality. Policies emerge from educational practice and social experience, and textual semantics are continuously revised through experiential feedback (Y. Liu et al., 2019; Shi & Zhang, 2012).

Implications and Conclusion

Drawing on intertextuality theory and applying the LDA topic model, this study conducts a thematic analysis of 23 science education policy documents from China and the United States. By examining the policies across three dimensions—topic structure, topic prominence, and topic evolution—the study reveals the underlying differences in the policy orientations of the two countries. The theoretical contributions of this study are threefold. First, it extends intertextuality theory from its original domain of literary analysis to the field of education policy analysis. By adopting an internal textual perspective, it uncovers divergences in the governance logic of science education between China and the United States. Second, whereas previous LDA-based studies primarily focused on classifying and ranking topics, often neglecting semantic linkages between policy texts and their embedded cultural or social contexts, this study goes beyond mere thematic categorization. It offers a theoretically grounded interpretation of the intertextual dynamics in Chinese and American science education policies. Third, the study enriches comparative education policy methodology. Conventional comparative policy research predominantly relies on qualitative approaches, such as content comparison or case study analysis. In contrast, this study demonstrates the value of integrating quantitative and qualitative methods within comparative policy inquiry, offering a more comprehensive analytical paradigm.

In terms of practical implications, three recommendations are proposed from the perspective of intertextuality theory. First, a multi-actor intertextual policy network should be established. Policymaking in science education ought to foster collaborative mechanisms among governments, schools, research institutions, and civil society organizations. Through semantic interconnection and reproduction, a multi-level and cross-sectoral governance system can be built to prevent policy fragmentation and semantic discontinuity. Second, the central role of teachers in science education must be strengthened. Teachers should be regarded as key nodes in the generation of policy semantics. Beyond addressing teacher quantity and training, policymakers should construct teacher empowerment mechanisms that enable teachers to engage in curriculum development, research collaboration, and social innovation initiatives. Third, semantic inheritance and cross-domain fusion in science education policy should be promoted. Policy language must remain both open and dynamic, allowing intertextual interaction with emerging domains such as artificial intelligence, sustainable development, and ecological civilization. At the same time, semantic continuity should be preserved to ensure that the science education ecosystem maintains a balance between stability and innovation.

Research Deficiencies

However, this study still has limitations. First, all selected samples are national-level policy texts, which limits the explanatory power of the findings for state- or provincial-level policies. In practice, local governments often formulate their own policy documents based on regional conditions, which introduces a certain degree of variation. Second, although this study adopts intertextuality theory as an analytical framework, its application remains exploratory. Future research could further integrate intertextuality theory with other education policy analysis frameworks to broaden theoretical perspectives and enhance the depth and breadth of policy interpretation.

Footnotes

Appendix A

Table A1.

Literature on Primary and Secondary Science Education Policy (China).

Number	Issuing Agency	Policy Document Name	Publication Date
1	国务院(The State Council of China)	《国家中长期科学和技术发展规划纲要(2006━2020年)》(Outline of the National Medium- and Long-Term Program for Science and Technology Development (2006–2020))	2006
2	教育部、科技部等(Ministry of Education, Ministry of Science and Technology, and others)	《建立中小学科普教育社会实践基地开展科普教育》(Establishing Social Practice Bases for Science Popularization in Primary and Secondary Schools to Promote Science Education)	2011
3	教育部(Ministry of Education)	《小学科学教育课程标准》(Primary School Science Education Curriculum Standards)	2017
4	教育部(Ministry of Education)	《教育部科学技术委员会章程》(Regulations of the Science and Technology Committee of the Ministry of Education)	2020
5	教育部(Ministry of Education)	《加强和改进中小学实验教学的意见》(Opinions on Strengthening and Improving Laboratory Teaching in Primary and Secondary Schools)	2019
6	国务院(The State Council of China)	《全民科学素质行动规划纲要(2021—2035年)》(Outline of the National Action Plan for Scientific Literacy (2021–2035))	2021
7	教育部等(Ministry of Education)	《利用科普资源助推“双减”工作的通知》(Notice on Using Science Popularization Resources to Promote the ‘Double Reduction’ Work)	2021
8	教育部(Ministry of Education)	《加强小学科学教师培养的通知》(Notice on Strengthening the Training of Primary School Science Teachers)	2022
9	国务院(The State Council of China)	《新时代进一步加强科学技术普及工作的意见》(Opinions on Further Strengthening Science and Technology Popularization in the New Era)	2022
10	科技部等(Ministry of Science and Technology, etc.)	《“十四五”国家科学技术普及发展规划》(14th Five-Year Plan for National Science and Technology Popularization Development)	2022
11	教育部等十八部门(Ministry of Education and 17 other departments)	《加强新时代中小学科学教育工作的意见》(Opinions on Strengthening Science Education Work in Primary and Secondary Schools in the New Era)	2023
12	教育部(Ministry of Education)	《基础教育课程教学改革深化行动方案》(Action Plan for Deepening the Reform of Basic Education Curriculum and Teaching)	2023
13	教育部(Ministry of Education)	《推荐首批全国中小学科学教育实验区、实验校的通知》(Notice on the Selection of the First Batch of National Experimental Zones and Experimental Schools for Science Education in Primary and Secondary Schools)	2023

Appendix B

Table B1.

Literature on Primary and Secondary Science Education Policy (United States).

Number	Issuing agency	Policy document name	Publication date
1	United States Congress	National Science and Technology Policy, Organization, and Priorities Act of1976	1976
2	United States Congress	America COMPETES Reauthorization Act of 2010	2010
3	National Research Council	Next Generation Science Standards: For States, By States	2013
4	United States Congress	STEM Education Act of 2015	2015
5	U.S. Department of Education	STEM 2026: A Vision for Innovation in STEM Education	2016
6	National Science and Technology Council	Charting a Course for Success America’s Strategy	2018
7	United States Congress	United States Innovation and Competition Act of 2021	2021
8	U.S. Department of Education	Raise The Bar:STEM Excellence for All Students	2022

ORCID iD

A. Y. M. Atiquil Islam

Ethical Considerations

Ethical approval was not required as the study did not involve human participants.

Author Contributions

Xin Li: conceptualization, methodology, investigation, writing—original draft. Tianhui Xu and A.Y.M Atiquil Islam: data collection and analysis, writing—original draft preparation, supervision, funding acquisition. Ping Zou: data collection, methodology. Jonathan Michael Spector: investigation, writing—reviewing and editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Peak Discipline Construction Project of Education at East China Normal University and Fundamental Research Funds for the Central Universities (2020ECNU-HLYT035).

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author* on reasonable request.

References

Aikenhead

G. S.

(1997). Student views on the influence of culture on science. International Journal of Science Education, 19(4), 419–428. https://doi.org/10.1080/0950069970190405

Allen

(2011). Intertextuality. Routledge.

Bianchi

Giorcelli

(2020). Scientific education and innovation: From technical diplomas to university STEM degrees. Journal of the European Economic Association, 18(5), 2608–2646.

Bezemer

Cowan

(2022). Exploring reading in social semiotics: Theory and methods (pp. 107–118). Routledge.

Blei

D. M.

A. Y.

Jordan

M. I.

(2003). Latent Dirichlet allocation. Journal of Machine Learning Research, 3, 993–1022.

Chen

Ren

Zheng

Yang

Cao

(2021). A comparison study of educational scientific collaboration in China and the USA. Physica A: Statistical Mechanics and Its Applications, 585, Article 126330. https://doi.org/10.1016/j.physa.2021.126330

Chen

Yao

Yang

(2016). Short text classification based on LDA topic model. In 2016 international conference on audio, language and image processing (ICALIP), Shanghai, China (pp. 749–753). IEEE.

Cheng

M. H. M.

Wan

Z. H.

(2015). Unpacking the paradox of Chinese science learners: Insights from research into Asian Chinese school students’ attitudes towards learning science, science learning strategies, and scientific epistemological views. Studies in Science Education, 51(2), 157–188.

Cong

Ding

Jia

Zeng

(2021). The influence of gender role on gender segregation of STEM majors in Chinese universities. In 2021 5th international seminar on education, management and social sciences (ISEMSS 2021), Chengdu, China (pp. 962–967). Atlantis Press.

10.

Deng

(2024). “Big science education” needs synergistic efforts from family, school, and society. Guangming Daily, 13, 1. https://link.cnki.net/doi/10.28273/n.cnki.ngmrb.2024.003001. [In Chinese]

11.

Foley

B. J.

Reveles

J. M.

(2014). Pedagogy for the connected science classroom: Computer supported collaborative science and the next generation science standards. Contemporary Issues in Technology and Teacher Education, 14(4), 401–418.

12.

Gan

(2021). Selection of the optimal number of topics for LDA topic model—Taking patent policy analysis as an example. Entropy, 23(10), 1301.

13.

García

Weiss

(2019). The teacher shortage is real, large and growing, and worse than we thought. The First Report in “The Perfect Storm in the Teacher Labor Market” Series. Economic Policy Institute.

14.

Goldhaber

Brown

Marcuson

Theobald

(2022). School district job postings and staffing challenges throughout the second school year during the COVID-19 pandemic (Working paper no. 273-1022). National Center for Analysis of Longitudinal Data in Education Research (CALDER).

15.

Huang

Zhao

(2015). Quantitative research on minority bilingual education policy texts in China from the Perspective of Policy Tools. Tsinghua University Journal of Education Research, 36(5), 88–95. https://link.cnki.net/doi/10.14138/j.1001-4519.2015.05.008808 [In Chinese]

16.

Jelodar

Wang

Yuan

Feng

Jiang

Zhao

(2019). Latent Dirichlet allocation (LDA) and topic modeling: Models, applications, a survey. Multimedia Tools and Applications, 78, 15169–15211.

17.

Johansen

J. D.

Larsen

S. E.

(2005). Signs in use: An introduction to semiotics. Routledge.

18.

Johnson

C. C.

(2012). Implementation of STEM education policy: Challenges, progress, and lessons learned. School Science and Mathematics, 112(1), 45–55. https://doi.org/10.1111/j.1949-8594.2011.00109.x

19.

Knain

Fredlund

Furberg

(2021). Exploring student reasoning and representation construction in school science through the lenses of social se miotics and interaction analysis. Research in Science Education, 51(1), 93–111.

20.

Learning Policy Institute. (2025, July 16). An overview of teacher shortages: 2025 [Polic report]. https://learningpolicyinstitute.org/product/overview-teacher-shortages-2025-factsheet

21.

Liang

L. L.

Liu

Fulmer

G. W.

(Eds.). (2017). Chinese science education in the 21st century: Policy, practice, and research. Springer.

22.

(2025). The historical evolution, internal logic, and reflective insights of governance decentralization in New York City public schools. Comparative Education Review, 47(4), 32–46. [In Chinese]

23.

Liu

Wang

(2015). Pre-service science teacher preparation in China: Challenges and promises. Journal of Science Teacher Education, 26(1), 29–44. https://doi.org/10.1007/s10972-014-9404-1

24.

Liu

(2023). STEM education in Japan, the United States, and China. Journal of Education, Humanities and Social Sciences, 13, 297–302. https://doi.org/10.54097/ehss.v13i.7917

25.

Liu

Tong

Zhao

Gao

Jiao

(2019). Analysis of science popularization policies issued by Chinese Central Government Organs and People’s Organizations from 2011 to 2015. Cultures of Science, 2, 109–127. https://doi.org/10.1177/209660831900200204

26.

Marginson

(2018). National/global synergy in the development of higher education and science in China since 1978. Frontiers of Education in China, 13, 486–512. https://doi.org/10.1007/s11516-018-0027-8

27.

Ming

(2023). Research on gender differences and gender discrimination in Chinese education. Lecture Notes in Education Psychology and Public Media, 31, 97–101. https://doi.org/10.54254/2753-7048/31/20231855

28.

Musschenga

Gosling

(2005). Science education and ethical values. Georgetown University Press.

29.

Nguyen

T. D.

Lam

C. B.

Bruno

(2022). Is there a national teacher shortage? A systematic examination of reports of teacher shortages in the United States (Vol. 22, pp. 1–78). Annenberg Institute at Brown University.

30.

Nöth

(1990). Handbook of semiotics. Indiana University Press.

31.

Oechsler

Borba

M. C.

(2020). Mathematical videos, social semiotics, and the changing classroom. ZDM, 52(5), 989–1001.

32.

Omenn

G. S.

(2006). Grand challenges and great opportunities in science, technology, and public policy. Science, 314(5806), 1696–1704. https://doi.org/10.1093/spp/14.4.182

33.

Qiao

(2025). The characteristics of the era, practical dilemmas, and promotion strategies of science education in primary and secondary schools. Curriculum, Teaching Material and Method, 45(7), 128–135. https://doi.org/10.19877/j.cnki.kcjcjf.2025.07.015 [In Chinese]

34.

Raj

P. P. E.

(2015). Text/texts: Interrogating Julia Kristeva’s concept of intertextuality. Ars Artium, 3, 77–80.

35.

Salvatierra

Cabello

V. M.

(2022). Starting at home: What does the literature indicate about parental involvement in early childhood science education?. Education Sciences, 12(3), 218.

36.

Samoyault

(2003). Intertextuality studies ( Wei

, trans.). Tianjin People’s Publishing House. [In Chinese]

37.

Sansone

(2017). Why does teacher gender matter?. Economics of Education Review, 61, 9–18.

38.

Scholes

(1982). Semiotics and interpretation. Yale University Press.

39.

Sha

(2020). Research on “famous school” construction under the perspective of intertextuality theory. Journal of Shaanxi Normal University (Philosophy and Social Sciences Edition, 49(3), 170–176. https://doi/10.15983/j.cnki.sxss.2020.0533 [In Chinese]

40.

Shamos

(1995). The myth of scientific literacy. Rutgers University Press.

41.

Shi

Zhang

(2012). Policy perspective on science popularization in China. In Schiele

Claessens

Shi

(Eds.), Science communication in the world (pp. 81–94). Springer. https://doi.org/10.1007/978-94-007-4279-6_5

42.

Smith

P. S.

Trygstad

P. J.

Banilower

E. R.

(2016). Widening the gap: Unequal distribution of resources for K-12 science instruction. Education Policy Analysis Archives, 24(8), n8.

43.

Staus

N. L.

O'Connell

Storksdieck

(2021). Addressing the ceiling effect when assessing STEM out-of-school time experiences. Frontiers in education, 6, Article 690431. https://doi.org/10.3389/feduc.2021.690431

44.

(2014). Introduction to public science and technology policy. Science Press. [In Chinese]

45.

Tan

(2015). Education policy borrowing and cultural scripts for teaching in China. Comparative Education, 51(2), 196–211. https://doi.org/10.1080/03050068.2014.966485

46.

Tao

Oliver

Venville

(2013). A comparison of approaches to the teaching and learning of science in Chinese and Australian elementary classrooms: Cultural and socioeconomic complexities. Journal of Research in Science Teaching, 50(1), 33–61. https://doi.org/10.1002/TEA.21064

47.

Wang

Zhang

Fang

Zhang

Zhao

Jiang

Qin

Yuan

Huang

Kang

(2022). Science education: A strategic frontier in great power competition—Research on international science education strategies and development paths. Journal of Chinese Education, (10), 25–31. [In Chinese]

48.

Worton

Still

(Eds.). (1990). Intertextuality: Theories and practices. Manchester University Press.

49.

(2020). Technology, power, and uncontrolled great power strategic competition between China and the United States. China International Strategy Review, 2(1), 99–119.

50.

Shi

Zhang

(2024). Evolution and development: Content and thematic analysis of central and local government open data management policies in China. Journal of Information Science, 42(4), 79–88. https://doi/10.13833/j.issn.1007-7634.2024.04.010 [In Chinese]

51.

Yang

Chen

(2024). Global comparison of STEM education. In Liu

N. C.

Feng

Wang

(Eds.), Education in China and the world: Achievements and contemporary issues (pp. 389–443). Springer.

52.

Yang

(2023). Policy text analysis in social science research: Methodology and methods. Social Science Journal, 12, 5–15. https://link.cnki.net/doi/10.13644/j.cnki.cn31-1112.2023.12.011 [In Chinese]

53.

Yao

J. X.

Guo

Y. Y.

(2018). Core competences and scientific literacy: The recent reform of the school science curriculum in China. International Journal of Science Education, 40(15), 1913–1933. https://doi.org/10.1080/09500693.2018.1514544

54.

Zhao

(2019). Is a new cold war inevitable? Chinese perspectives on US-China strategic competition. The Chinese Journal of International Politics, 12(3), 371–394. https://doi.org/10.1093/CJIP/POZ010

55.

Zhu

(2011). Educational policy research. Beijing Normal University Press. [In Chinese]