An Ai-Supported Biostatistics E-Course Based on the Successive Approximation Model: Evaluation in Medical Education

The Successive Approximation Model (SAM) provides an iterative framework for developing digital learning resources, while artificial intelligence (AI) may enhance personalization and cognitive support. However, empirical evidence on the combined implementation of SAM and AI in medical education remains limited. This study aimed to evaluate the acceptability and perceived educational effectiveness of a biostatistics e-course developed using the SAM model and supplemented with an AI-supported component.

A mixed-perspective descriptive study was conducted following the development and implementation of a modular biostatistics e-course based on the SAM at Karaganda Medical University. The course was developed through iterative prototyping, stakeholder feedback, and progressive refinement consistent with the SAM framework. Participants included 215 undergraduate medical students, 73 students involved in the needs assessment phase, and course developers. Questionnaires and descriptive analysis were used to gain staff perceptions of the SAM model of course development, student use of AI, and student evaluation of the quality of the course.

Across key evaluation domains, 64.7% of students reported positive perceptions of the course, including improved understanding of biostatistics and greater convenience compared with traditional learning formats, while negative responses remained below 12%. The AI component was used by 56.9% of students, primarily for explanation of theoretical material and analysis of statistical concepts. Among AI users, 82.1% reported improved understanding, and 53.7% reported increased motivation. However, 57.7% encountered errors, and trust in AI remained moderate. Developers positively evaluated the SAM model, particularly its iterative design and flexibility, while highlighting the need for methodological training and institutional support.

The integration of SAM-based instructional design and AI-supported learning represents a feasible and acceptable approach to developing adaptive digital courses in medical education. Effective implementation depends on the balanced integration of instructional design, AI-mediated support, and institutional readiness.