Analysis of learning behavior characteristics based on process evaluation data

Abstract

Strengthening process evaluation is an important topic in the current educational evaluation reform. Applying process evaluation to the current popular blended teaching will help promote the process of educational evaluation reform. In this paper, a process evaluation index system for blended teaching is constructed based on the Delphi method and the AHP. The evaluation data are collected in the actual course, and the K-means clustering algorithm is used to mine and analyze the characteristics of students' learning behavior under the blended teaching mode. It is found that under the process evaluation model, students can be divided into five types of learners: independent research type, confident and flexible type, digital ability type, cooperative and mutual aid type, and learning difficulty type. The main learning behaviors are learning engagement, classroom performance, online learning, assignment submission, unit assessment and problem solving. The polarization is severe, and the comprehensive ability of high-level skills is poor. Accordingly, the author puts suggestions to provide a practical reference and theoretical basis for front-line teachers to carry out the process evaluation of blended teaching.