A1 Journal article (refereed), original research
Just Passing Courses or Learning: Building Bayesian Classifier to Assess Outcomes
Open Access publication
Publication Details Authors: Kasurinen Jussi, Nikula Uolevi
Publication year: 2016
Language: English
Related Journal or Series Information: International journal on information technologies and security Volume number: 8
Issue number: 4
Start page: 3
End page: 14
JUFO-Level of this publication: 1
Open Access: Open Access publication
|
Abstract
It is widely known that learning programming is difficult, and there are studies which confirm this truth. But even if the basic measurements such pass rates or average grades are acceptable, are the students actually learning the programming constructs, or just passing the course requirements by chance? In this paper we report our results from creation of a knowledge-based classifier system based on the Bayesian Knowledge Tracing algorithm. Our results suggest that the passing rates and average grades do not tell the whole story; even if the course results indicate only small changes, the acquired programming knowledge may vary drastically.