Predicting Quitting in Students in Speed-Reading Course: A Case Study

Bildiriler > Özet Bildiri · 2023

Yazarlar: İnanç KABASAKAL

The 10th InTraders International Conference on Social Sciences and Education, Makedonya, Üsküp, Uluslararası, Sözlü Sunum, Özet Bildiri, 2023-2023, ss 10-10

Özet

<p>Speed-reading&nbsp;courses&nbsp;are&nbsp;designed&nbsp;to&nbsp;improve&nbsp;their&nbsp;students’&nbsp;reading&nbsp;speed&nbsp;and&nbsp;comprehension.&nbsp;The&nbsp;use&nbsp;of&nbsp;e-learning&nbsp;environments&nbsp;enables&nbsp;data&nbsp;collection&nbsp;that&nbsp;helps&nbsp;assessment&nbsp;of&nbsp;students&nbsp;and&nbsp;predictive&nbsp;analyses.&nbsp;Student&nbsp;dropout&nbsp;prediction&nbsp;is&nbsp;among&nbsp;the&nbsp;popular&nbsp;problems&nbsp;in&nbsp;this&nbsp;context.&nbsp;This&nbsp;study&nbsp;presents&nbsp;a&nbsp;neural&nbsp;network-based&nbsp;prediction&nbsp;model&nbsp;that&nbsp;identifies&nbsp;impending&nbsp;dropouts&nbsp;from&nbsp;a&nbsp;speed-reading&nbsp;course.&nbsp;Assessment&nbsp;scores&nbsp;obtained&nbsp;in&nbsp;the&nbsp;last&nbsp;ten&nbsp;sections&nbsp;were&nbsp;analyzed&nbsp;to&nbsp;predict&nbsp;dropouts&nbsp;who&nbsp;will&nbsp;not&nbsp;proceed&nbsp;to&nbsp;the&nbsp;next&nbsp;level.&nbsp;Despite&nbsp;the&nbsp;challenge&nbsp;of&nbsp;predicting&nbsp;short-term&nbsp;dropouts,&nbsp;the&nbsp;tests&nbsp;resulted&nbsp;in&nbsp;an&nbsp;accuracy&nbsp;of&nbsp;81%&nbsp;on&nbsp;average,&nbsp;while&nbsp;57%&nbsp;of&nbsp;predicted&nbsp;students&nbsp;actually&nbsp;dropped&nbsp;out&nbsp;before&nbsp;the&nbsp;next&nbsp;level,&nbsp;and&nbsp;53%&nbsp;of&nbsp;dropouts&nbsp;were&nbsp;successfully&nbsp;predicted&nbsp;on&nbsp;time.</p>

Detay görüntüle