장르별·분야별 이독성 예측 변인에 기반한 독서지수 개발 연구

초록

This study calculated a reading index using multiple regression analysis by exploring variables that predict readability by genre and academic field. The findings revealed that in the literary genre, the key predictors were the 'total number of tokens', 'ratio of first-grade vocabulary', and 'primary lexical density'. In the humanities and social sciences, the predictors included the 'total number of content words', 'vocabulary grade', 'ratio of modifiers', and 'number of connective endings per sentence'. For the scientific field, the 'total number of content words', 'ratio of second-grade vocabulary', and 'MATTR of content words' were selected as the predictors of readability. In addition, the reading index developed in this study demonstrated an explanatory power ranging from 55.8% to 69.8%. The specific equations are as follows: • (reading grade in the literary genre) = (0.0562) * (total number of tokens/1000) + (-39.9131) * (primary lexical density) + (-12.0132) * (the ratio of first-grade vocabulary) + 30.8099 • (reading grade in the humanities and social genres) = (0.0564) * (total number of content words/1000) + (5.9177) * (vocabulary grade) + (69.1019) * (the ratio of modifier) + (0.9015) * (the number of connective endings per sentence) - 8.0712 • (reading grade in science genre) = (0.0508) * (total number of content words/1000) + (39.7540) * (the ratio of second-grade vocabulary) + (21.4726) * (MATTR of content words) - 15.4678 The variation in predictive variables across genres and fields underscores the necessity for tailored reading indices. Furthermore, intensified research into the development of reading indices by genre and field is expected to significantly enhance our understanding of genre characteristics and disciplinary literacy.

키워드