STUDENT-GENERATED QUESTIONS AS FORMATIVE PEDAGOGICAL SIGNALS: A FUZZY-APRIORI INTERPRETIVE MODEL IN PRIMARY SCHOOL SETTINGS
Éva Karl1*, György Molnár2*
1PhD Student, Doctoral School of Multidisciplinary Engineering Sciences (MMTDI), Széchenyi István University, Győr, Hungary
2Full Professor, Head of Teacher Training of Engineering Education Institution, Kandó Kálmán Faculty of Electrical Engineering, Óbuda University, Budapest, Hungary
∗ Correspondence: molnar.gyorgy@uni-obuda.hu
DOI : 10.66538/DH.2026.2.1.
Abstract
This study investigates how end-of-lesson student-generated questions can be interpreted as formative pedagogical signals. Conventional performance indicators - tests and grades - typically capture the endpoint of learning rather than its process. We present an algorithmic model (WTCAi) grounded in fuzzy set theory and Apriori-based association rule mining, capable of jointly processing the cognitive depth and content of student questions, as well as learner background variables. The empirical dataset comprises data from 202 students and 174 parents, including 5,832 student-generated questions collected in grades 7-8 across digital culture, science, and physics lessons. Fuzzy representation accommodates partial truths and the inherent uncertainty of classroom data, while the Apriori algorithm yields interpretable, teacher-readable rule families. Key associative patterns identified include: students who formulate high-abstraction questions despite low parental support tend to perform above expectations, and lessons employing cooperative or visually rich teaching methods co-occur with significantly more complex student questions. The model also incorporates a dynamic rule-updating mechanism that cyclically integrates new student data. The findings carry both data-mining novelty and direct pedagogical relevance: they support differentiated planning and timely scaffolding within an inclusive, formative assessment framework. The study is exploratory; subsequent phases will focus on formal reliability procedures and longitudinal validation across multiple school settings.
Keywords: formative assessment; student-generated questions; fuzzy sets; association rule mining; educational data mining; learning analytics; inclusive pedagogy; interpretability; adaptive learning; pedagogical decision support.