Abstract


In order to validate the usability of a newly developed ¡Çconversational CAI¡Ç, three kinds of laboratory experiment have been sequentially conducted. The users are 40 university students in the first and second experiments, and 3 students in the third experiment. And they compare the conversational CAI with conventional CAI of just ¡Çtext-display-and-see¡Ç type but with more rich contents than those by the conversational CAI. The first step was performed by achievement tests to the forty users by dividing them into two groups of learning through either conversational CAI or conventional CAI in order to verify statistically the effectiveness of the conversational CAI by self learning way. The second step was the questionnaire survey to all the users who experienced the both types of CAI to confirm the effect of various features implemented in the conversational CAI. The last step was the application of psycho-physiological measurement to a few users who used the conversational CAI, in order to estimate dynamic changes of their arousal level and thinking level and to find out the parts to be modified for further improvement of the conversational CAI designing.

Powered by JDBS Ver.2.1.0
(C) 2020 Hirotake Ishii