Further Results from the Evaluation of an Intelligent Computer Tutor to Coach Self-Explanation
ITS '00 Proceedings of the 5th International Conference on Intelligent Tutoring Systems
| Hi-index | 0.05 |
Self-explanation is an instructional strategy that has shown to be beneficial for math and science learning. However, it remains an open question whether these benefits will extend to other domains like second language grammar learning. Within the domain of the English article system (teaching students when to use a, an, the, or no article at all), we compare two computer-based tutoring conditions in an in vivo classroom study. In the article choice condition, students select the correct article to complete the sentence. In the explanation choice condition, students are given a sentence with the correct article highlighted and choose the rule or feature that best explains the article use. Students (N=101) in both conditions show significant learning on both procedural (article choice) and declarative (explanation choice) tasks. Not surprisingly, we found that declarative instruction (explanation choice) led to significant learning of explanations, while procedural practice (article choice) led to significant learning of the procedures. More interestingly, we also found evidence of cross-type transfer such that declarative practice led to procedural gains and procedural practice led to better understanding of the declarative rules. In general the effects of prompted self-explanation appeared somewhat stronger than those of procedural practice.