Search

本研究ではベトナム人日本語学習者92名を対象に、漢字の学習方法と学習意識の質問紙調査および漢字能力テストを実施した。探索的因子分析で学習方法と学習意識の質問紙調査の結果を解析し、ベトナム人日本語学習者による漢字学習における特有の方法と意識を解明した。そして、学習方法では(1)グループ化、(2)母語活用、(3)練習の3つの因子、学習意識では(1)情意性、(2)困難性、(3)有効性の3つの因子を抽出した。さらに、学習方法および学習意識と漢字能力との因果関係を構造方程式モデリングの統計手法で検討した。その結果、ベトナム人日本語学習者は、漢字に対する意識が学習方法に影響し、さらに学習方法が漢字能力へ促進的に貢献するという逐次的な因果関係を持つことを示した。本研究では、ベトナム人日本語学習者の漢字に対する学習方法および学習意識を具体的に描き、どのようなアプローチが漢字習得に貢献するのかを明らかにした。, Two questionaire surveys on study of kanji method (learning methods and learning consciousness) and two kanji proficiency test (writing and reading tests) were conducted on 92 Vietnamese learners of Japanese. Exploratory factor analysis was applied to clarified the learning methods and consciousness of learners, which gave 3 groups. The 3 methods which Vietnamese learners of Japanese used the most are “Grouping the same kanji”, “Use mother language to remember the meaning and the pronunciation of kanji” and “Writing practice of kanji”. The 3 conscious learning which learners of kanji are “The affection for the use of kanji”, “The difficulty of studying kanji” and “The effectiveness of using kanji”. Furthermore, the cause-and-effect relationship between learning methods, learning consciousness and kanji proficiency was examined using structural equation modeling. It was found that Vietnamese learners of Japanese consciousness of kanji influence their learning, and moreover that the learning-method has sequential causal relationship contributing to the kanji proficiency. In this study, we concretely described the study of kanji among Vietnamese learners of Japanese, and clarified what kind of the approach contributes acquisition of kanji proficiency.

Most of the analytical methods of factor rotation have been developed to achieve the simple structure of multiple factors. Rotation methods for the bi-factor structure constructed of a general factor and group factors have also been proposed. Monte Carlo experiments are performed to study of the behavior of factor rotation methods for recovering of the bi-factor structure. The condition of this study is the rotation method (Varimax, Promax, Bi-factor, Biquartimin, and Quartimin). Three types of factor correlations are also examined as the part of population models. Biquartimin method reveals better recovering of the bi-factor structure among above analytic rotation methods. The implications of these findings relating the factor structure and factor rotation are discussed.