Does framing a problem-solving scenario influence learning?
In this paper, we discuss potential effects of framing, a pedagogical strategy used by some teachers, with the view of implementing it in an intelligent tutoring system. The process of framing a learning activity, in our case problem solving, consists of having the activity in between a pre-action (or priming) phase and a post-action (or reflective) phase. We also describe an evaluation study where the experimental and control groups participated in a framed and non-framed learning session respectively. The pre- and post-action phases were whiteboard group sessions led by an SQL domain expert. The problem-solving phase was done solely on SQL-Tutor, an ITS for database querying. During the problem-solving phase, the experimental group solved the same number of problems in 23% less time than their counterparts in the control group. The type of problems solved and the high-level help used were similar for both groups. Although the learning gains were high, they were similar across groups, questioning the effectiveness of group-based reflection. The experimental group was also more efficient, expending less effort for similar gains. After examining various learning theories and analyzing the study results, we conclude that this is a valid teaching format for the next step of our research: investigating its implementation and evaluation within an ITS.