Dictation : その理論的背景・評価としての妥当性・誤答の類型

抄録

The purpose of this paper is to investigate dictation from the following viewpoints : the theoretical background, validity as an integrative testing device, and the classification of errors. First, a survey of studies and research on dictation is presented. There have been various views and reports concerning dictation, some of which are favorable, and some other few unfavorable. Since the end of the 1960's, however, dictation has attracted the attention of many scholars as an integrative testing device as well as an overall-skill teaching device. Our specific interest herein is to research dictation from the viewpoint of an integrative teaching and/or testing device. To validate the assumption that dictation can be an integrative testing device, an experiment was made. The aim of the experiment was to seek correlation between dictation and communicative competence, that is, listening comprehension ability and speaking ability. The results obtained showed a relatively high correlation betweeen dictation and these abilities, with a correlation coefficient of .703-.692. These data are high enough to ascertain that dictation is a reliable and valid testing device to measure EFL learners' communicative competence integratively. As for the classification of errors, we proposed a system of classification as follows: errors are first classified into three categories, that is, no response, inaccuracies in listening, and spelling mistakes. Furthermore, inaccuracies in listening are divided into four categories, that is, phonological, morphological, syntactic, and semantic errors. By applying the errors made by EFL learners to these classification categories, their difficulties and their general tendencies of errors will be clarified. The results can be utilized as a feedback for a better learning and teaching of EFL.