Scholarlink

This study aims to utilize big data technologies to analyze the learning behavior patterns and differences of students of varying achievement levels in online courses on the “Tailored Education Network” offered by the Ministry of Education in Taiwan. The data for this study was sourced from the learning records of 72 students on the “Tailored Education Network” platform between June and August 2023, totaling 101,730 entries. Each data entry included user ID, checkpoint response results, video interaction sequence number, video-watching behavior, and exercise accuracy rates. The data underwent preprocessing for data consolidation, the removal of entries with missing or duplicate values, and the coding and definition of video-watching behaviors. After preprocessing, data from 48 students with complete records remained, totaling 5857 entries. Subsequently, cluster analysis was performed using the K-means method based on two indicators of learning effectiveness: Checkpoint response results and exercise accuracy rates. This analysis grouped students into high (27), medium (13), and low (8) achievement categories. Lag sequential analysis was applied to each group’s data to interpret the frequency and relationship of behavioral transitions and provide information on the behavioral sequences or learning behavior patterns of the students across different achievement levels. Students of high achievement frequently reviewed videos to conduct checkpoint evaluations and take notes, adjust video speed, switch screen modes, and check the video browsing page. Students of middle achievement adjusted playback speed, reviewed content, paused videos, and switched to full screen for better focus. Students of low achievement moderately paused, occasionally rewound, adjusted playback settings, and sporadically visited the video browsing page post-viewing. Students of high achievement were more engaged in checkpoint quizzes for self-assessment and reviewed videos or paused for note-taking upon encountering doubts, which was less observed by students of middle or low achievement.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Applying Big Data Analysis to Learning Behavior Patterns and Their Differences Among Diverse Achievers in Online Courses

  • 会议论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年

查看全文
  • Yi Chen,
  • Chang-Tai Li

摘要

This study aims to utilize big data technologies to analyze the learning behavior patterns and differences of students of varying achievement levels in online courses on the “Tailored Education Network” offered by the Ministry of Education in Taiwan. The data for this study was sourced from the learning records of 72 students on the “Tailored Education Network” platform between June and August 2023, totaling 101,730 entries. Each data entry included user ID, checkpoint response results, video interaction sequence number, video-watching behavior, and exercise accuracy rates. The data underwent preprocessing for data consolidation, the removal of entries with missing or duplicate values, and the coding and definition of video-watching behaviors. After preprocessing, data from 48 students with complete records remained, totaling 5857 entries. Subsequently, cluster analysis was performed using the K-means method based on two indicators of learning effectiveness: Checkpoint response results and exercise accuracy rates. This analysis grouped students into high (27), medium (13), and low (8) achievement categories. Lag sequential analysis was applied to each group’s data to interpret the frequency and relationship of behavioral transitions and provide information on the behavioral sequences or learning behavior patterns of the students across different achievement levels. Students of high achievement frequently reviewed videos to conduct checkpoint evaluations and take notes, adjust video speed, switch screen modes, and check the video browsing page. Students of middle achievement adjusted playback speed, reviewed content, paused videos, and switched to full screen for better focus. Students of low achievement moderately paused, occasionally rewound, adjusted playback settings, and sporadically visited the video browsing page post-viewing. Students of high achievement were more engaged in checkpoint quizzes for self-assessment and reviewed videos or paused for note-taking upon encountering doubts, which was less observed by students of middle or low achievement.