Hybrid Ensemble Learning Model for Measuring the Creativity of Iraqi Gifted Students

Main Article Content

Zainab Khyioon Abdalrdha
Rawaa waleed Abd Al wahab
Saad Ahmed Dheyab
Rawaa Ibrahim Esa

Abstract

Creativity is the defining trait of giftedness and an essential skill in the age of artificial intelligence (AI). This research presents a hybrid ensemble-learning model to differentiate the creativity profiles of gifted human students and an AI. Six domains of creativity (fluency, flexibility, originality, elaboration, sensitivity, and redefinition) were visualized as the unique features used to build 12,000 data points (24,000 samples, 12,000 responses from gifted human students, and the rest from the generative AI) with identical prompts and completed as the responses to a creativity scale instrument administered to Iraqi gifted students. Six classifiers (decision tree, random forest, K-nearest neighbor, support vector machines, ANN, and hybrid ensemble model with an eXtreme Gradient Boosting) XGBoost( meta-learner) were tested, and the hybrid ensemble showed the best test accuracy (96.75%), followed by SVM (96.73%).


The cross-validation stability was higher with SVM (mean = 96.69%, SD = 0.0014). MANOVA showed statistically significant multivariate behavior (p < 0.001), with gifted students showing significant superiority against the AI in all six creativity aspects. Feature importance weightings showed sensitivity (0.414) and elaboration (0. 191) are the strongest classifiers in differentiating human and AI responses, while discriminating effect sizes for sensitivity (1.464) and originality (1.359) show major differences. AI responses were rated equally high to gifted students' responses in fluency and elaboration but ranked lower in originality, sensitivity, and redefinition. Bottom line, this research is an able and explainable model for scale-based differentiation between human and AI creativity

Article Details

How to Cite
Hybrid Ensemble Learning Model for Measuring the Creativity of Iraqi Gifted Students. (2026). Journal of the College of Basic Education, 2(عدد خاص), 1355-1379. https://doi.org/10.35950/cbej.v2iعدد خاص.16247
Section
pure science articles

How to Cite

Hybrid Ensemble Learning Model for Measuring the Creativity of Iraqi Gifted Students. (2026). Journal of the College of Basic Education, 2(عدد خاص), 1355-1379. https://doi.org/10.35950/cbej.v2iعدد خاص.16247