Motivational Design of Support for Learning Effective Problem-Selection Strategies in an Intelligent Tutoring System

• In Event: Self-Regulated Learning Analytics: Aligning Data and Their Treatment to the Assumptions of Theory

Abstract

Objective & Theoretical Framework. Many online learning technologies grant students great autonomy and control, which impose high demands for self-regulated learning (SRL) skills. With the fast development of online learning technologies, helping students acquire SRL skills becomes critical to student learning. The current work focuses on supporting students’ learning of a central SRL skill, making effective problem selection decisions in online learning environments with the aid of certain kinds of learning analytics that are commonly available in many learning technologies. Research has shown that especially younger learners are poor at selecting problems strategically based on their learning status (how much has been learned for different learning units, as generally displayed by the learning analytics). Prior studies mainly targeted supporting students’ problem selection in systems where the scaffolding was in effect, but few studies have tried to teach students the transferrable skills that can be applied in new learning environments.
Our design centers on teaching an effective problem selection rule, namely, to select problem types that are not fully mastered (“Mastery Rule”), through motivational design in an intelligent tutoring system (ITS).

Methods. We have designed interventions such as explicit instruction, feedback messages, and rewards through a user-centered design process. The interventions aim at helping students want to apply the Mastery Rule when they are given control over problem selection, in addition to helping them correctly apply the rule. We have conducted a classroom experiment with around 300 6th – 8th grade students to investigate the effectiveness of the interventions on supporting students’ learning of the Mastery Rule in an ITS for equation solving, Lynnette. There were two phases in the experiment, the learning phase and the transfer phase. In the transfer phase, the scaffolding on problem selection was removed, which allows us to investigate whether the students would be able to apply the Mastery Rule without the scaffolding. Figure 1 shows the problem selection screen of Lynnette. As shown in Figure 1, the problem selection screen displays the student’s progress towards mastery for the levels, calculated by Bayesian Knowledge Tracing. The students are free to select any problem levels they want to work on.

Data analysis. We will present analyses of the tutor log data to observe students problem selection – whether the students had selected any mastered problems in the tutor during the learning and transfer phases. We will also incorporate data from paper pre–, mid–, and posttests to measure students’ learning of equation solving and self-reported enjoyment. This work contributes to the research of supporting SRL in ITSs through a motivational design perspective. It also allows investigation of the transfer of SRL skills with objective behavioral measures from system log data.

Authors

• Yanjin Long, University of Pittsburgh

• Vincent Aleven, Carnegie Mellon University