Predictive Analytics of Math Anxiety in Students: A Machine Learning Study on PISA 2022 Turkey Data

Yıl 2025, Cilt: 1 Sayı: 1, 1 - 14, 31.03.2025

Büşra Yağcı , Murat Şahin , Zehra Akbiyik , Yunus Doğan

Öz

Mathematics anxiety is the worry, fear, and stress individuals experience in mathematics-related situations. Mathematics anxiety is an important problem in the education system and an important factor affecting students' academic success. In this context, studies to prevent or reduce mathematics anxiety are of great importance. Machine learning algorithms significantly contribute to such studies by enabling the extraction of information from large data sets. PISA 2022 dataset focuses on the assessment of student performance in mathematics, reading and science to measure the extent to which students can use what they learned in and out of schools for their full participation in societies. Some 690 000 students took the assessment in 2022, representing about 29 million 15-year-olds in the schools of the 81 participating countries and economies. The primary purpose of this study is to predict mathematics anxiety of students in Turkey using the PISA 2022 dataset. So, the dataset has been filtered based on Turkey. The new dataset includes 7250 instances and 1280 feature attributes. In order to use this dataset, a multi-stage preprocessing is carried out. Two different datasets are developed by selecting different attributes. In Dataset A, there are 26 attributes and 6065 instances. The current study also generated another dataset including attributes containing PISA weighted scores which is called Dataset B. Variables with weighted averages of the PISA 2022 data set were used in feature selection for Dataset B. Mathematics anxiety values in both datasets are calculated using Decision Tree (DT), Random Forest (RF), Ada Boost (AB), Gaussian Naive Bayes (GaussianNB), K Nearest Neighbors (KNN), Multi-Layer Perceptron Classifier (MLPC), and XGBoost (XGB). These models are compared to calculating Precision, Recall, F1-Score, and Accuracy values.

Anahtar Kelimeler

PISA 2022, mathematics anxiety, machine learning algorithms, data mining

Öğrencilerde Matematik Kaygısının Tahmini Analitiği: PISA 2022 Türkiye Verileri Üzerine Bir Makine Öğrenmesi Çalışması

Öz

Matematik kaygısı, bireylerin matematikle ilgili durumlarda deneyimledikleri endişe, korku ve strestir. Matematik kaygısı, eğitim sisteminde önemli bir sorundur ve öğrencilerin akademik başarısını etkileyen önemli bir faktördür. Bu bağlamda, matematik kaygısını önleme veya azaltma çalışmaları büyük önem taşımaktadır. Makine öğrenimi algoritmaları, büyük veri kümelerinden bilgi çıkarılmasını sağlayarak bu tür çalışmalara önemli ölçüde katkıda bulunmaktadır. PISA 2022 veri seti, öğrencilerin okullarda ve okul dışında öğrendiklerini toplumlara tam katılımları için ne ölçüde kullanabildiklerini ölçmek için matematik, okuma ve fen alanlarındaki öğrenci performansının değerlendirilmesine odaklanmaktadır. 2022'de yaklaşık 690.000 öğrenci değerlendirmeye katıldı ve bu, 81 katılımcı ülke ve ekonominin okullarındaki yaklaşık 29 milyon 15 yaşındaki öğrenciyi temsil ediyor. Bu çalışmanın temel amacı, PISA 2022 veri setini kullanarak Türkiye'deki öğrencilerin matematik kaygısını tahmin etmektir. Bu nedenle, veri seti Türkiye bazında filtrelenmiştir. Yeni veri seti 7250 örnek ve 1280 özellik niteliği içermektedir. Bu veri setini kullanabilmek için çok aşamalı bir ön işleme gerçekleştirilir. Farklı nitelikler seçilerek iki ayrı veri seti oluşturulur. Veri Seti A'da 26 nitelik ve 6065 örnek bulunur. Mevcut çalışmada ayrıca PISA ağırlıklı puanları içeren nitelikler içeren Veri Seti B adı verilen başka bir veri seti de üretilmiştir. PISA 2022 veri setinin ağırlıklı ortalamalarına sahip değişkenler, Veri Seti B için özellik seçiminde kullanılmıştır. Her iki veri setindeki matematik kaygısı değerleri Karar Ağacı (DT), Rastgele Orman (RF), Ada Boost (AB), Gauss Naive Bayes (GaussianNB), K-En Yakın Komşu (KNN), Çok Katmanlı Algılayıcı Sınıflandırıcı (MLPC) ve XGBoost (XGB) kullanılarak hesaplanmıştır. Bu modeller Kesinlik, Duyarlılık, F1 Puanı ve Doğruluk değerlerinin hesaplanmasıyla karşılaştırılmıştır.

Anahtar Kelimeler

PISA 2022, matematik anksiyetesi, makine öğrenmesi algoritmaları, veri madenciliği

Kaynakça

• Acıslı Celik, S., & Yesilkanat, C. M. (2023). Predicting science achievement scores with machine learning algorithms: a case study of OECD PISA 2015–2018 data. Neural Computing and Applications, 35(28), 21201-21228.
• Araújo, L., & Costa, P. (2023). Reading to Young Children: Higher Home Frequency Associated with Higher Educational Achievement in PIRLS and PISA. Education Sciences, 13(12), 1240. https://doi.org/10.3390/educsci13121240
• Arpa, T., & Çavur, M. (2024). A Comparative Analysis of Machine Learning Techniques to Explore Factors Affecting Mathematics Success in Developing Countries: Turkey, Mexico, Thailand, And Bulgaria Case Studies (Doctoral dissertation, M. Hanefi CALP).
• Bayirli, E. G., Kaygun, A., & Öz, E. (2023). An Analysis of PISA 2018 Mathematics Assessment for Asia-Pacific Countries Using Educational Data Mining. Mathematics, 11(6), 1318. https://doi.org/10.3390/math11061318
• Bernardo, A. B., Cordel, M. O., Lucas, R. I. G., Teves, J. M. M., Yap, S. A., & Chua, U. C. (2021). Using machine learning approaches to explore non-cognitive variables influencing reading proficiency in English among Filipino learners. Education Sciences, 11(10), 628.
• Bernardo, A. B. I., Cordel, M. O., II, Lapinid, M. R. C., Teves, J. M. M., Yap, S. A., & Chua, U. C. (2022). Contrasting Profiles of Low-Performing Mathematics Students in Public and Private Schools in the Philippines: Insights from Machine Learning. Journal of Intelligence, 10(3), 61. https://doi.org/10.3390/jintelligence10030061
• Bi, Z. J., Han, Y. Q., Huang, C. Q., & Wang, M. (2019). Gaussian naive Bayesian data classification model based on clustering algorithm. In 2019 International Conference on Modeling, Analysis, Simulation Technologies and Applications (MASTA 2019), 396-400.
• Boreham, I. D., & Schutte, N. S. (2023). The relationship between purpose in life and depression and anxiety: A meta‐analysis. Journal of Clinical Psychology, 79(12), 2736-2767.
• Breiman, L. (2001). Random forests. Machine learning, 45, 5-32.
• Chen, T., & Guestrin, C. (2016). Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, 785-794.
• Dai, S., Hao, T., Ardasheva, Y., Ramazan, O., Danielson, R. W., & Austin, B. (2023). PISA reading achievement: Identifying predictors and examining model generalizability for multilingual students. Reading and Writing, 36(10), 2763-2795.
• Dong, X., & Hu, J. (2019). An exploration of impact factors influencing students’ reading literacy in Singapore with machine learning approaches. International Journal of English Linguistics, 9(5), 52-65.
• Freund, Y., & Schapire, R. E. (1996). Experiments with a new boosting algorithm, 96, 148-156.
• Gabriel, F., Signolet, J., & Westwell, M. (2018). A machine learning approach to investigating the effects of mathematics dispositions on mathematical literacy. International Journal of Research & Method in Education, 41(3), 306-327.
• Guo, G., Wang, H., Bell, D., Bi, Y., & Greer, K. (2003). KNN model-based approach in classification. In On The Move to Meaningful Internet Systems 2003: CoopIS, DOA, and ODBASE: OTM Confederated International Conferences, CoopIS, DOA, and ODBASE 2003, 986-996.
• Haw, J. Y., & King, R. B. (2023). Understanding Filipino students’ achievement in PISA: The roles of personal characteristics, proximal processes, and social contexts. Social Psychology of Education, 26(4), 1089-1126.
• Hernández-Ramos, J. P., & Martínez-Abad, F. (2023). Professional Development among Secondary Teachers in Spain: Key Associated Factors as of PISA 2018. Journal of Intelligence, 11(5), 93. https://doi.org/10.3390/jintelligence11050093
• Khine, M. S., Liu, Y., Pallipuram, V. K., & Afari, E. (2024). A Machine-Learning Approach to Predicting the Achievement of Australian Students Using School Climate; Learner Characteristics; and Economic, Social, and Cultural Status. Education Sciences, 14(12), 1350. https://doi.org/10.3390/educsci14121350
• Lee, H. (2022). What drives the performance of Chinese urban and rural secondary schools: A machine learning approach using PISA 2018. Cities, 123, 103609.
• Lezhnina, O., & Kismihók, G. (2022). Combining statistical and machine learning methods to explore German students’ attitudes towards ICT in PISA. International Journal of Research & Method in Education, 45(2), 180-199.
• Li, Z., & Li, Q. (2024). How Social Support Affects Resilience in Disadvantaged Students: The Chain-Mediating Roles of School Belonging and Emotional Experience. Behavioral Sciences, 14(2), 114. https://doi.org/10.3390/bs14020114
• Liu, A., Wei, Y., Xiu, Q., Yao, H., & Liu, J. (2023). How Learning Time Allocation Make Sense on Secondary School Students’ Academic Performance: A Chinese Evidence Based on PISA 2018. Behavioral Sciences, 13(3), 237. https://doi.org/10.3390/bs13030237
• Luo, S. (2023). Factors Affecting English Reading in Macao, Hong Kong, and Singapore: Combining Machine Learning Methods and Hierarchical Linear Regressions Using PISA 2018 Data (Doctoral dissertation, University of Macau).
• Masci, C., Johnes, G., & Agasisti, T. (2018). Student and school performance across countries: A machine learning approach. European Journal of Operational Research, 269(3), 1072-1085.
• Merriam-Webster Dictionary. (n.d). Anxiety. Retrieved May 1, 2024, from https://www.merriam-webster.com/dictionary/anxiety
• Pejić, A., Molcer, P. S., & Gulači, K. (2021). Math proficiency prediction in computer-based international large-scale assessments using a multi-class machine learning model. In 2021 IEEE 19th International Symposium on Intelligent Systems and Informatics (SISY) (pp. 49-54). IEEE.
• Puah, S. (2020). Predicting students’ academic performance: a comparison between traditional MLR and machine learning methods with PISA 2015.
• Ramazan, O., Dai, S., Danielson, R. W., Ardasheva, Y., Hao, T., & Austin, B. W. (2023). Students’ 2018 PISA reading self-concept: Identifying predictors and examining model generalizability for emergent bilinguals. Journal of School Psychology, 101, 101254.
• Rebai, S., Yahia, F. B., & Essid, H. (2020). A graphically based machine learning approach to predict secondary schools performance in Tunisia. Socio-Economic Planning Sciences, 70, 100724.
• Sirganci, G. (2023). A Machine Learning Approach to Assess Differential Item Functioning of PISA 2018 ICT Engagement Questionnaire: Item Functioning of PISA 2018 ICT Engagement Questionnaire. International Journal of Curriculum and Instruction, 15(3), 2079-2093.
• Tzora, V. A. (2025). Defining the Predictors of Financial Literacy for High-School Students. Journal of Risk and Financial Management, 18(2), 45. https://doi.org/10.3390/jrfm18020045
• Von Lorenz, A. C. (2025). Exploring Latent Class Profiles of Mathematics Performance: Insights from PISA 2022 Using Growth Mindset Indicators and Group Comparison Analysis. Journal Evaluation in Education (JEE), 6(1), 150-158.
• Zheng, J. Q., Cheung, K. C., & Sit, P. S. (2024). Identifying key features of resilient students in digital reading: Insights from a machine learning approach. Education and Information Technologies, 29(2), 2277-2301.