摘要
本文通过对口语测试常用的三种评分标准——分解评分、整体评分和任务分项评分——的同质性分析以及效度、信度和可操作性程度的比较研究发现:这三种评分标准在评分结果上无实质区别,分解评分标准的三个构成要素实属同一个因素。分解评分标准是经过严格培训的授权考官使用效度和信度最高的标准,而整体评分标准是经过初步培训的非授权考官使用效度和信度最佳的标准。考虑到大学英语口语考试的实际情况,本文认为整体评分标准是最符合大学英语大面积口试实际、操作性较强、效度和信度相对较高的评分标准。
By comparing the validity,reliability and operationalization of the three rating scales of oral English test:analytic scoring,holistic scoring and item analytical scoring,the authors find that there is no great difference in the scores rated with the three scales and the three factors comprising the analytic rating scale in fact belong to one factor; and that to CET-SET raters,the analytic rating scale has the highest validity and reliability,whereas to non-CET-SET raters,the holistic rating scale is the most valid and reliable one. Taking into account the reality of college oral English test,the authors hold that holistic rating scale,because of its ease to operate and comparatively high validity and reliability to non-CET-SET raters,is the rating scale most suitable for large-scale college oral English test.
出处
《外语教学与研究》
CSSCI
北大核心
2008年第6期440-446,共7页
Foreign Language Teaching and Research