Towards Adaptive Worked-Out Examples In An Intelligent Tutoring System

Worked-out examples (WOEs) have been shown to be effective for learning, but they need to be adapted to student characteristics. We experimented with three versions of our Intelligent Tutoring System for Computer Science, one that does not include WOEs, and two that differ as concerns WOE length and content. We found that shorter WOEs are more effective for advanced students, whereas novice students learned the same, no matter the presence or length of the WOE.