Research on the Model of Differentiated Instruction Supported by Learning Behavior Analysis——Taking the Course of Probability and Statistics as an Example
Differentiated instruction respects learners'internal cognitive logic,and provides suitable learning environments and learning interventions by tapping and diagnosing learners'individual potentials and strengths.However,the existing differentiated instruction has problems such as poor timeliness and single dimension,which remains at the shallow level of intervention.Data driven learning behavior analysis provides a new way to design and implement high-level differentiated intervention strategies.Based on the learning log data of 30 college students registered in the innovation class of Probability and Statistics on the Chaoxing learning platform,an empirical study on online learning behavior characteristics and personalized intervention was conducted using correlation analysis,clustering analysis,and lag sequential analysis.According to the observation results,out of the 62 behavioral sequences formed by the 10 defined learning behaviors,7 behavioral sequences were significantly correlated to the final test score.Using the frequencies of 62 behavioral sequences as classification variables,30 students were divided into three categories based on their learning behavior patterns using k-means algorithm.Three learning styles were discovered and identified using the lag sequence analysis method,including mixed/disordered pattern,isolated/decomposed pattern,and holistic/planned pattern,and a differentiated intervention framework as well as specific intervention strategies were proposed.This study is helpful for constructing a technology-enhanced differentiated instruction model and providing a new paradigm for building and optimizing personalized student-centered classrooms.
differentiated instructionlearning behavior analysisProbability and Statisticsdata drivenintervention strategy
万方数据
