高校生の自由英作文はどのように評価されているのか -分析的評価尺度と総合的評価尺度の比較を通しての検討- How are high school students’ free compositions evaluated by teachers and teacher candidates? :A comparative analysis between analytic and holistic rating scales

Page No.: 
189
Writer(s): 
山西博人(Yamanishi Hiroyuki) 広島大学

The aim of this paper was to clarify the characteristics of the evaluation of high school students’ free compositions by high school teachers and university students who are teacher-candidates. This was done by analyzing a comparison of the results of analytic and holistic rating scales. In addition, information about their consciousness regarding the evaluation and instruction of free compositions was obtained with the help of a questionnaire. The pedagogical implications were explored with the help of the analyses and the questionnaire.
In recent years, free compositions in English (for example, explaining a particular situation and arguing for or against a statement) have been increasingly required in university entrance examinations in Japan. Such kinds of activities are also required in situations of actual communication. Moreover, the renewed course of study (Ministry of Education, 1999) has been conducted in high schools since the academic year of 2003. The overall objective of the course of study is to develop students’ practical communication abilities such as expressing their own ideas through written English. To satisfy these requirements, the chances of instructing students on free compositions in high school classrooms will be increased. In such a situation, it is important to obtain information on how teachers and teacher-candidates instruct and evaluate students’ free compositions.
The survey consisted of several analyses of the teachers’ and teacher-candidates’ evaluation data, and a questionnaire concerning their consciousness of the evaluation and instruction of free compositions. The raters who evaluated high school students’ free compositions were 10 high school teachers from a national high school and two public high schools and 6 university students who belonged to a teacher-training course in a national university and had experience with practical teaching.
First, the raters evaluated 40 compositions written by 20 high school students who belonged to the first and the second grade of a national high school in the Chugoku region in Japan (all the 20 students wrote two kinds of compositions). The evaluations were carried out using holistic rating scales and two kinds of analytic rating scales. In this survey, two subscales given by Ishida and Mori (1985), that is, the objective holistic rating scale (How good do you think the composition is?) and the subjective holistic rating scale (How do you personally like the composition?) were adopted as holistic rating scales. At the same time, two kinds of rating scales, that is, Jacobs, Zinkgraf, Wormuth, Hartfiel, and Hughey’s (1981) ESL Composition Profile and National Institute for Educational Policy Research’s (2002) kantenbetsu-hyoka (it means an analytic rating), were adopted as analytic rating scales. The following two points were discussed on the basis of a comparison of the result of the evaluation of the holistic and analytic rating scales using descriptive statistics, Pearson’s product-moment correlation coefficients, and Cronbach’s alpha coefficients: 1) To clarify the characteristics of the evaluations performed by the teachers and teacher-candidates, and 2) To examine the reliability and the validity of the kantenbetsu-hyoka scale. From the results, it appeared that the teachers’ evaluations were consistent within and beyond the rating scales, while those performed by the university students who were teacher-candidates were slightly inconsistent within and beyond the rating scales. At the same time, it also appeared that the validity of the kantenbetsu-hyoka scale was not higher than that of the other scales.
Second, an open-ended questionnaire, which included two parameters, was given to the same raters. The first question was what they regarded as being of greatest importance while instructing students on free compositions, and the second was what they regarded as being of greatest importance when they evaluated them. The following point was discussed on the basis of the questionnaire: 3) To obtain information on the consciousness of teachers and teacher-candidates concerning the evaluation of and instruction on free compositions. From the results of the questionnaire, it appeared that teachers were more sensitive to the conditions of high school students than the university students who were teacher-candidates.
These results led to the pedagogical implications as follows: 1) The validity of the kantenbetsu-hyoka scale was not higher than the other scales; therefore, it would be better to use it in combination with other scales, 2) It would be better for teachers to carefully consider each parameter of the analytic rating scales, in order to make good use of the merits of the scale, and 3) It would be better for university students who are teacher-candidates to have more opportunities of evaluating high school students’ compositions and of instructing them on writing compositions, before they become teachers.
本研究では、複数の評価尺度を組み合わせることで、教員と大学生による高校生の自由英作文評価の実態が調査された。総合的評価尺度として「客観的総合評価」と「主観的総合評価」が、分析的評価尺度としてESL Composition Profileと「観点別評価」が使用された。調査では、10名の教員と6名の大学生が、20名の高校生によって書かれた40編の自由英作文を評価した。評価結果を記述統計量、信頼性係数、相関係数から分析することで、教員の評価は大学生の評価よりも一貫性が高いことが示された。また「観点別評価」の信頼性と妥当性も議論され、妥当性がやや低かったことが示された。同時に、自由記述の質問紙による指導と評価の意識調査も行われ、教員は大学生よりも指導現場に根ざした意識を有していることが確認された。これらの結果に対する考察が行われ、そこから教員と大学生の双方に対する教育的示唆が示された。

PDF: