Abstract
Meta-cognition is close to reflective thinking that is aware thinking of its own conceptual and contextual as well as being conscious of the reasons and evidence that support hypothesis or conclusion. The influence and relevance of meta-cognition in learning and problem solving have been extensively demonstrated, and aptitude may even be more important for problem solving success. The purposes of this study are to understand meta-cognitive ability of Japanese students and to study the good assessment methodologies. Research subjects are 74 high school first grade students and 40 high school second grade students. Artificial Neural Network (ANN) is used to determined strategy of students in solve problems. All of groups are 34 groups, but 24 groups who could solve the problem. The results show that percentages of Japanese High School students are 75.0% in low category, 20.8% in intermediate category, and 4.2% in high category. Multiple choice problems are not suggested for student assessment, because many students were used guessing to solve the problem, where guessing is not thinking. After students engaged in the assessment, they felt their motivation increasing to deep understanding in learning chemistry.
