Search

近年、中国では、「一人っ子政策」の継続や、IT 技術の進展による間接的なコミュニケーションの増加などの社会事情により、若者の社会的スキルが懸念されるようになった。そこで、若者の社会的スキルを測定し、問題を発見・解決することが急務となる。若者の社会的スキルの測定にあたって、これまで、日本でも、欧米でも多くの社会的スキル尺度が開発されてきたが、中国においては、この種の尺度はまだ開発されていない。そこで、本研究では、中国の若者に適用できる社会的スキル尺度を開発する試みとして、中国と同じ集団文化に属する日本で開発された、若者の社会的スキルを測定する「KiSS-18(Kikuchi’s Social SkillScale ・ 18 項目版)」という尺度が中国で適用できるかどうかを検討した。中国の高校生を対象として検討を行った結果、尺度全体の信頼性がある程度確認できた一方、下位尺度の信頼性は必ずしも高くなかった。また、尺度全体の平均得点、および因子構造に関して、日本における先行研究との間に違いが存在していた。今後、中国文化、中国人の価値観および社会的行動などの特徴を考慮した検討が必要であろう。, A lot of social skill scales were developed in the U.S. and in Japan. However, there were obtained few in China. Before developing an original social skill scale for the Chinese, we would examine whether the Japanese social skill scale named KiSS-18 can be suitable to Chinese teenagers or not because a lot of similarities exist between Japan and China. Factor analysis was done by using 741 students of senior high school in China, and it was found that the factor structure was different from the composition of the KiSS-18 scale in Japan. In spite of high reliability that we confirmed through the whole scale, those of subscales were not so high. Moreover, average score through the whole scale of the Chinese students is much higher than their Japanese partners. These results suggested that it is uncertain when apply a Japanese social skill scale to Chinese. To increase the validity and reliability, it is necessary to develop an original social skill scale based on Chinese culture and values in the future.

The back-propagation (BP) network is widely recognized as a powerful training tool of the multilayer neural networks (MLNNs). Usually it suffers from a slow convergence rate and often results in local minimums, since it applies the steepest descent method to update the network weights. A variety of related algorithms have been introduced to address that problem. Levenberg-Marquardt algorithm is one of the fastest types of these algorithms. This paper presents an approximation calculation for Hesse matrix to train the neural networks when the second order item can not be omitted, and the improved algorithm is successfully used in the surface flaws evaluation of the conductive materials based on eddy current testing (ECT).