Availability and Effectiveness of Generative AI for Web-Based Investigative Learning

Abstract

In Web-based learning, learners are expected to investigate information through gathering and navigating Web resources, and to construct knowledge. In our previous work, we designed a model of Web-based investigative learning, which represents a learning process as a cycle of phases. We also developed a cognitive tool called iLSB (interactive Learning Scenario Builder) for scaffolding learning process as modeled. At present, on the other hand, learners can explore information by chatting with generative Al, which generates texts in a dialogue format. However, it is unclear about the context where generative Al is available for Web-based investigative learning, and about the effective use of generative Al. In order to examine the availability and effectiveness of generative Al, we have had case studies including the comparison between the use of iLSB and the use of generative Al. The results suggest that generative Al prevents learners from investigating sufficiently compared with iLSB, and the Web-based investigative learning model contributes to the proper use of generative Al for investigative learning. Furthermore, it is suggested that the use of generative Al is effective in exploring background knowledge about questions investigated and in organizing information for knowledge construction.

Ähnliche Arbeiten