EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Why AI-based adaptive learning outperforms game-based strategies in the development of mathematical reasoning

Vera Septi Andrini, Umi Hidayati, Erdyna Dwi Etika, Irfan Yusuf, Saida Ulfa

Frontiers in Education · 22.05.2026 · 11

DOI: 10.3389/feduc.2026.1812582 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Frontiers in Education, online erschienen am 22.05.2026. Verfasst von Vera Septi Andrini und 4 weiteren Autor:innen. Die Arbeit nennt 86 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Mathematical reasoning remains the most challenging dimension of the international mathematics literacy framework. As digital transformation reshapes educational practice, AI-based adaptive learning and gamification have emerged as two widely adopted instructional approaches. However, empirical evidence comparing their differential associations with the development of mathematical reasoning remains limited, particularly for pre-service mathematics teachers. This study examines the relationship between AI-based adaptive learning and mathematical reasoning, as well as gamification strategies and mathematical reasoning, among pre-service mathematics teachers. Employing an explanatory quantitative design, the study involved 110 pre-service mathematics teachers from a private university in East Java. Data were collected using a five-point Likert scale questionnaire measuring AI-based adaptive learning (X1; 3 indicators, 12 items) and gamification strategies (X2; 2 indicators, 10 items), alongside open-ended cognitive task questions (Q23–Q28) scored with an analytical rubric (0–4 per item; maximum total score=24) to assess mathematical reasoning (Y; 2 indicators, 6 items). Data analysis utilized Partial Least Squares Structural Equation Modelling (PLS-SEM) with 5,000 bootstrapping resamples, evaluating measurement model validity and structural model path coefficients. The HTMT ratio between AI-based adaptive learning and mathematical reasoning (0.6822) fell below the 0.85 threshold, confirming adequate discriminant validity; thus H1 is supported. Conversely, gamification strategy showed no statistically significant relationship with mathematical reasoning ( β = −0.065; t = 0.856; p = 0.392; f² = 0.006), meaning H2 is not supported. The model achieved an Adjusted R² of 0.4561 and R² of 0.4463, explaining 45.61% of the variance in mathematical reasoning with moderate predictive relevance (Q² = 0.259). PLS predict RMSE (0.842) was lower than the LM-RMSE (0.860), indicating superior predictive accuracy over a linear baseline. These findings suggest that mathematics teacher education programs should prioritize AI-based adaptive technologies that emphasize reasoning processes, while reorienting gamification strategies toward justification, reflection, and evaluation of mathematical solutions.

Zitieren

APA

Andrini, V. S., Hidayati, U., Etika, E. D., Yusuf, I., & Ulfa, S. (2026). Why AI-based adaptive learning outperforms game-based strategies in the development of mathematical reasoning. Frontiers in Education, 11. https://doi.org/10.3389/feduc.2026.1812582

Vancouver

Andrini VS, Hidayati U, Etika ED, Yusuf I, Ulfa S. Why AI-based adaptive learning outperforms game-based strategies in the development of mathematical reasoning. Frontiers in Education. 2026;11. doi:10.3389/feduc.2026.1812582

BibTeX
@article{Andrini2026,
  author = {Andrini, Vera Septi and Hidayati, Umi and Etika, Erdyna Dwi and Yusuf, Irfan and Ulfa, Saida},
  title = {Why AI-based adaptive learning outperforms game-based strategies in the development of mathematical reasoning},
  journal = {Frontiers in Education},
  year = {2026},
  volume = {11},
  doi = {10.3389/feduc.2026.1812582}
}
Kostenlos starten →