Analysis of the Characteristics of Mathematical Ability that Affect the Discrimination and Prediction of JMO Qualifiers

Abstract

　　In this study, I analyzed the mathematical characteristics that contribute to the discrimination and prediction of successful applicants by conducting the following two studies.

　　　(1) Logistic regression analysis was conducted for each group of students according to their academic ability and installer, and the characteristics that contribute to the discrimination and prediction of successful applicants were estimated.

　　　(2) Logistic regression analysis was conducted for all general high school students to supplement the above conjecture.

　　The following five mathematical characteristics were identified as a result of the study.

　　　・Interest in mathematical analysis

　　　・Interest in mathematical problem solving

　　　・Logical thinking ・Visualization and summary of arguments and results

　　　・Understanding of the main points of problems and the methods of their solutions

　　For the top four, the stronger the feature, the more it is expected to contribute to the discrimination and prediction of successful candidates, and for the bottom feature, the stronger the feature, the lower the probability of discrimination and prediction.

　　In Tamura (2018), the“ creativity” factor was extracted in a factor analysis conducted on successful candidates. On the other hand, Sternberg (1988) divided thinking styles into five categories, all of which are necessary for creativity, and it is inferred that creativity has a role in connecting between item attributes. The five characteristics mentioned in this study straddle all four attributes of the questionnaire items: ① interest, motivation, and attitude toward mathematics, ② mathematical thinking, ③ skill and expression in thinking, and ④ knowledge and understanding of mathematics.

　　This result is suggested to complement the claim of Sternberg (1988).