How much Learning Support Should be Provided to Novices and Advanced Students?

Learning from examples, either alone or combined with problem solving has been proven to be beneficial for learning in Intelligent Tutoring System. However, it is generally unknown how much example-based assistance should be provided. We previously found that erroneous examples prepared students better for problem solving in comparison to worked examples when the order of learning activities is fixed [2]. However, students do not necessarily need all learning activities. We introduced a novel strategy which adaptively decides which learning activity (a worked example, an incorrect example, a problem, or none at all) is appropriate for a student based on his/her performance in SQL-Tutor. In this paper, we investigate the effect of the adaptive strategy on students with different levels of prior knowledge. We found both novices and advanced students who received learning activities adaptively achieved the same learning outcomes as their peers in a fixed condition, but with fewer learning activities. Surprisingly, there was no significant difference on the number of learning activities between novices and advanced students.