The REDEEM authoring environment allows teachers to create simple ITSs from existing CBT by imposing their pedagogical preferences about how different groups of students should best be taught. Consequently, a fundamental assumption behind the research is that students will learn either more effectively or more efficiently from these ITSs than from the original CBT. We conducted two studies with 14-16 yr old students learning genetics to test this assertion. In both experiments, a class teacher who had expert knowledge of both the topic and of the students constructed two ITSs with REDEEM from pre-existing CBT. Using a crossover design, the learning outcomes for students who studied these two courses (either a REDEEM then CBT course or vice versa) were compared. In the first study, we found that performance of the students (N = 86) improved from pre-test to post-test but learning outcomes were not influenced by type of learning environment. However, inspection of the process data revealed that students who engaged with REDEEM’s features did learn more. In a second study, conducted in a more naturalistic context, a further 15 students completed the courses The results of this study revealed that REDEEM could significantly improve learning compared to CBT. Detailed analysis of participants’ performance suggested that REDEEM could enhance knowledge by supporting additional interactivity but that features such as macro-adaptation did not appear to impact upon performance. Possible interpretations of these results are discussed in the light of the many evaluation issues for ITS authoring tools.
Download the pdf here