A study on comparative advantages of concept map construction and self-explanation

Abstract

Being knowledgeable means not only knowing the concepts, ideas, terms, and rules that make up the knowledge domain, but also understanding their relationships correctly. Concept maps have been found effective in representing meaningful relationships between the concepts and enhancing students’ learning activities. In this research, we test whether concept mapping can be further improved by incorporating self-explanation activities. Specifically, while controlling for total training time, we compared three alternative learning conditions: control group (no concept mapping), regular concept mapping group (provided with an expert skeleton diagram), and concept mapping reinforced with self-explanation group (started with an expert skeleton and finished with self-explanation activities), in the experiment setting of learning a new programming language, Ruby.

Undergraduate and graduate students at a higher education institute in Korea, KAIST, were invited to participate in the experiment. Before the experiment, the participants took a pre-test and were randomly assigned to one of the three experimental conditions, resulting in 20 students in each group. Learning outcomes were measured for declaration knowledge (multiple choice questions) and skill compilation (near-transfer and far-transfer). Study results revealed significant differences (1) between the control group and the regular concept mapping group and (2) between the control group and the concept mapping reinforced with self-explanation group, for both declarative knowledge and far-transfer skill compilation. The group differences were more pro-nounced between the control group and the concept mapping with self-explanation group even though the differences between the alternative concept mapping conditions were not statistically different. The results overall indicate that the proposed method of concept mapping reinforced with self-explanation is the most effective among the three experimental conditions ...