Formative Assessment Practices and Academic Progress among School Students

Abstract

Formative assessment has evolved from periodic classroom checking into a potentially continuous, data-informed process capable of identifying learning difficulties before they become persistent achievement gaps. This study proposes an artificial intelligence–supported formative assessment framework for examining academic progress among school students through the joint analysis of assessment performance, response behaviour, feedback uptake, and longitudinal learning trajectories. The central research gap concerns the limited integration of pedagogically meaningful formative-assessment indicators with interpretable machine-learning models that can distinguish temporary performance fluctuations from sustained academic risk.
Rather than treating quiz scores as isolated measures, the proposed approach conceptualizes progress as a dynamic pattern involving mastery growth, misconception recurrence, response-time stability, feedback responsiveness, and assessment consistency.
A longitudinal student-learning dataset is proposed in which classroom formative-assessment events are transformed into temporal learner profiles and analysed through explainable machine-learning techniques.