1. Research Topic
This study investigates the factors enabling students to achieve both high academic performance (reading proficiency) and high learning well-being (life satisfaction and school belonging) in the context of Chinese education. Based on PISA 2018 data from four Chinese provinces/municipalities (Beijing, Shanghai, Jiangsu, Zhejiang), it explores how family, school, and individual characteristics predict the dual excellence of “academic success and happiness.”

2. Research Questions
Primary RQ:
What factors at the family, school, and individual levels predict students’ likelihood of becoming high academic performers (reading proficiency), and which additional factors further characterize high-performing students with elevated well-being (life satisfaction and school belonging)?

Sub-RQs:

How do socioeconomic status (ESCS), teacher qualifications, and metacognitive skills hierarchically distinguish high-performing students from their peers?

Beyond academic predictors, what role do emotional resources  and psychological traits play in differentiating "dual-excellence" students (high performance + high well-being) from high performers with low well-being?

How can these findings inform actionable strategies to cultivate both academic excellence and holistic well-being in competitive educational systems?

3. Objectives

Primary Objective:
To systematically identify (1) predictors of high academic performance and (2) distinguishing features of "dual-excellence" students through machine learning-driven analysis of PISA 2018 data.

Specific Objectives:

Quantify the relative importance of material resources, institutional factors , and cognitive skills in predicting academic success.

Map the interplay between emotional-environmental factors and student psychological traits in shaping dual excellence.

Propose context-sensitive strategies for policymakers and educators to foster "academically strong and emotionally thriving" students, with implications for global education systems facing similar challenges.

4. Theoretical Framework

Ecological Systems Theory (Bronfenbrenner): Examines interactions between family, school, and individual factors.

5. European/International Dimension

Cross-Cultural Relevance: While focused on China, the findings address a global challenge: balancing academic excellence with student well-being. PISA data ensures methodological alignment with international standards.

Policy Implications: Results inform debates on teacher training (e.g., advanced degrees vs. pedagogical enthusiasm), parental engagement strategies, and metacognitive skill development—issues critical to OECD countries facing similar educational pressures.

Comparative Value: The dual focus on academic and emotional outcomes complements European initiatives like the OECD’s “Future of Education and Skills 2030,” advocating holistic student development.

6. Innovation

Methodological: Combines machine learning (random forest/decision trees) with traditional educational research to uncover non-linear relationships among predictors.

Conceptual: Challenges the “either/or” dichotomy between academic performance and well-being, proposing a synergistic model for global education reform.

1. Data Source & Sample This study utilizes data from the Programme for International Student Assessment (PISA) 2018, focusing on 12,058 students aged 15 across 361 schools in Beijing, Shanghai, Jiangsu, and Zhejiang. The weighted sample represents 960,300 students in these provinces. Variables 2. Dependent Variables: High Academic Performance: Students scoring at PISA Reading Level 5 or above (top 24.8% in the sample). High Well-being: Students reporting life satisfaction ≥7/10 and no sense of alienation at school. Dual Excellence: Students meeting both criteria (11.51% of the sample). 3. Independent Variables: 46 features across three levels: Family:Socioeconomic status (ESCS), parental emotional support (EMOSUPS)…… School:Teacher qualifications(PROAT5AM,PROAT6), school climate (PERCOMP, TEACHINT)…… Individual: Metacognitive skills (METASPAM), psychological traits (EUDMO, GFOFAIL)…… 4. Analytical Approach Machine Learning Models: Random Forest: Identified key predictors via Mean Decrease Gini. Decision Trees: Visualized hierarchical interactions among predictors (e.g., METASPAM → ESCS → PROAT5AM). 5. Validation: Data split: 80% training, 20% testing. Performance metrics: Accuracy (0.65–0.83) and F1-score (0.70–0.89). Confusion matrix analysis ensured robustness. 6. International Comparability PISA’s standardized methodology (e.g., matrix sampling, IRT scaling) ensures cross-country validity. Machine learning bypasses cultural biases in traditional regression, enhancing generalizability.

This study reveals that cultivating “dual-excellence” students—those achieving both academic mastery and emotional well-being—requires a synergistic approach integrating material, institutional, and emotional resources. While socioeconomic status (ESCS) and teacher qualifications (e.g.,PROAT5AM) remain critical predictors of academic success, emotional support from parents (EMOSUPS) and pedagogical enthusiasm from teachers (TEACHINT) emerge as decisive factors distinguishing high-performing students with well-being from those struggling emotionally. Notably, students’ metacognitive strategies (e.g.,METASPAM) and psychological resilience（e.g.,EUDMO, RESILIENCE) mediate the impact of competitive school climates, suggesting that cognitive and emotional skills can buffer external pressures. These findings challenge the pervasive “zero-sum” narrative pitting academic rigor against student happiness. Instead, they advocate for policies that: 1) Empower teachers through advanced training and passion-driven professional development; 2) Reframe parental engagement from academic supervision to emotional scaffolding; 3) Integrate metacognitive and socio-emotional learning into curricula to foster self-regulated, resilient learners. The machine learning-driven methodology (random forest/decision trees) demonstrates its utility in disentangling complex educational ecosystems, offering a replicable framework for global contexts. By aligning with OECD’s holistic education goals, this study contributes actionable insights for balancing excellence and well-being in high-stakes systems worldwide.

Aldridge, J. M., Fraser, B. J., Fozdar, F., et al. (2016). Students’ perceptions of school climate as determinants of wellbeing, resilience and identity. Improving Schools, 19(1), 5–26. Artelt, C., & Schneider, W. (2015). Cross-country generalizability of the role of metacognitive knowledge in students’ strategy use and reading competence. Teachers College Record, 117(1), 1–32. https://doi.org/10.1177/016146811511700104 Beaman, R., Wheldall, K., & Kemp, C. (2006). Differential teacher attention to boys and girls in the Classroom. Educational Review, 58(3), 339–366. Christenson, S., Reschly, A. L., & Wylie, C. (2012). Handbook of research on student engagement. New York, NY: Springer. Gorostiaga, A., & Rojo-Álvarez, J. L. (2016). On the use of conventional and statistical-learning techniques for the analysis of PISA results in Spain. Neurocomputing, 171(1), 625-637. Hallinan, M. T. (2008). Teacher influences on students’ attachment to school. Sociology of Education, 81(3), 271-283. Hidayati, M., Inderawati, R., & Loeneto, B. (2020). The correlations among critical thinking skills, critical reading skills and reading comprehension. English Review: Journal of English Education, 9(1), 69–80. https://doi.org/10.25134/erjee.v9i1.3780 Kılıç-Depren, S., & Depren, Ö. (2021). Cross-cultural comparisons of the factors influencing the high reading achievement in Turkey and China: Evidence from PISA 2018. The Asia-Pacific Education Researcher, 31, 427–437. https://doi.org/10.1007/s40299-021-00584-8 Kotu, V., & Deshpande, B. (2018). Data science: Concepts and practice (2nd ed.). Elsevier. Leto, I. V., Petrenko, E. N., & Slobodskaya, H. R. (2019). Life satisfaction in Russian primary schoolchildren: Links with personality and family environment. Journal of Happiness Studies, 20(6), 1893-1912. Ma, X. (2003). Sense of belonging to school: Can schools make a difference? The Journal of Educational Research, 96(6), 340-349. OECD. (2009). Top of the class: High performers in science in PISA 2006. Paris: OECD Publishing. OECD. (2019a). PISA 2018 results: Where all students can succeed. Paris: OECD Publishing. OECD. (2019b). PISA 2018 Assessment and Analytical Framework. PISA, OECD Publishing, OECD. (2019c). PISA 2018 results: What students know and can do. Pairs: OECD Publishing. Sanzana, M. B., Garrido, S. S., & Poblete, C. M. (2015). Profiles of Chilean students according to academic performance in mathematics: An exploratory study using classification trees and random forests. Studies in Educational Evaluation, 44, 50-59. Schlatter, E., Molenaar, I., & Lazonder, A. W. (2020). Individual differences in children’s development of scientific reasoning through inquiry-based instruction: Who needs additional guidance? Frontiers in Psychology, 11, 1–14. https://doi.org/10.3389/fpsyg.2020.00904