The Online Course Tool for Adaptive Learning (OCTAL) is a tool that combines an exercise system with a concept map to allow learners to explore an underlying prerequisite structure of topics. An algorithm that estimates a learner's level of mastery highlights concepts in the graph to provide the user with a metacognitive hint about their progress through the material. Learners are guided by the prerequisite structure and knowledge inference but may navigate freely through the graph. We intend OCTAL to be a formative assessment tool that is not tied to any specific course or subject and provide authoring tools for content designers to create material. Toward the goal of being usable in a number of online courses, OCTAL has support to be embedded within online learning platforms such as edX. Students enroll in online courses with different learning goals and, as a result, may wish to pursue their own paths through the material. OCTAL presents the underlying prerequisite structure of the material to allow learners the opportunity to decide whether or not deviation from the expert-defined path would be beneficial for their understanding. This allows students to metacognitively consider their level of mastery in a course's advanced concepts by exploring exercises without limitation and may therefore be useful to help answer the question "will this course be useful?" Similarly, for those students enrolled in a course, it allows them to decide how to prioritize consumption of content and discover which concepts they may reasonably skip, if necessary. In order to study the benefits of metacognition with OCTAL, we authored a concept map and question set for topics from UC Berkeley's CS10: The Beauty and Joy of Computing. We presented the tool to the students of the course between their first and second midterm exams during the spring of 2014. In the study, we found no statistically significant changes in metacognition among participants who used the tool. However, analysis of participant usage of the tool reveals differences in the way learners approach concepts presented to them in a list versus in a graph. In particular, while users often followed a list of concepts in-order, learners that navigated a graph explored concepts in clusters.




Download Full History