Learning strategy feedback system for students using data mining technology
ABSTRACT: Using data mining technology, we developed a feedback system of learning strategies based on models that were created from data obtained from high school students. Using this system, we evaluated it from the following perspectives: (a) model validity, (b) effectiveness of the metaphors built into the system, and (c) users’ subjective evaluations. Results showed the following. (a) The models were validated because the difference of academic ability estimated from the models was reflected in data obtained from a monitor test. (b) The signal metaphor, representing the achievement levels of learning strategies, was effective because the length of time spent to read explanations of learning strategies differed according to the signal color; the direct road metaphor, representing the order of learning strategies, was understood by nearly half of the students. (c) Some users reported that the system animations were too long, but the overall system evaluation was favorable.
Presider: Miyazoe Terumi, Tokyo University of Science