Developing an Artificial Neural Network-based model for predicting EFL achievement level

• Published in: Multimedia tools and applications Vol. 84; no. 24; pp. 28061 - 28084
• Main Authors: Bouzayenne, Amani, Harizi, Riadh
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.07.2025; Springer Nature B.V
• Subjects: Artificial intelligence; Artificial neural networks; Cognitive style; College students; Colleges & universities; Computer Communication Networks; Computer Science; Data analysis; Data Structures and Information Theory; English as a second language learning; Foreign languages; Language acquisition; Learning; Learning strategies; Motivation; Multimedia Information Systems; Neural networks; Personality traits; Special Purpose and Application-Based Systems; Speech perception; Students; Track 5: Multimedia and Education; Artificial Neural Network; Learners’ differences; Predicting EFL achievement level
• ISSN: 1573-7721; 1380-7501; 1573-7721
• DOI: 10.1007/s11042-024-20295-8

Motivated by the increasing applications of artificial intelligence in education, the present study aimed at developing an Artificial Neural Network (ANN)-based model for predicting EFL (English as a Foreign Language) achievement levels using data about learners’ variables. Another endeavor of the current survey was to analyze the importance of learners’ differences in predicting EFL achievement: personality traits, perceptual learning style preferences, language learning strategies, motivation, and attitudes. A Likert-response format questionnaire served to gauge the considered learners’ variables. The subjects were 200 Tunisian first-year university students aged 19 to 23, not English-specialized students. As for data analysis procedures, the employment of the alpha internal consistency coefficient was to inform about the reliability of the study instrument, and the ANNs were to give insights into the interconnections among the study variables. The findings postulated that it was practicable to implement systematic ANNs to predict learners’ EFL achievement levels using their individual affective and cognitive factors, as the model achieved an excellent overall accuracy (an accuracy of 97%). Moreover, results proved that using learners’ variables simultaneously provided a better prediction of EFL achievement. The study highly recommends the implementation of ANNs for the analysis of EFL learning to increase its effectiveness.