摘要
口语测试是主观性较强的测试项目,测试结果与许多因素相关。本文将统计学原理以及模糊理论应用于英语口语测试,通过评分模板的引入以降低考官的上述因素对考生的实际成绩评定的影响。同时,评分模板的引入使评分更为精确,便于考试结果标准化,避免因考题太难或太容易而导致整批考生成绩偏高或偏低。本文报告了利用评分模板将考生的名次折算成成绩,并与面试成绩的比较实验,在此基础上建构学生的标准成绩。实验表明,该方法获得的成绩与传统的面试法获得的成绩有很好的一致性。
Spoken language testing is a subjective process, the results of which are frequently affected by many factors during testing. This paper offers a procedure to reduce subjectivity. Specifically, a scoring scheme is adopted, based on the principles of statistics and the fuzzy theory. By means of this scheme the ranked oral performance is statistically converted to scores, rendering evaluation more objective. Standardization prevents both overachievement and underachievement caused by excessively simple or difficult testing contents. An experiment was run to test the effectiveness of the scoring method. The results demonstrate a high degree of agreement between the new method and the traditional interview format.
出处
《现代外语》
CSSCI
北大核心
2004年第1期77-82,共6页
Modern Foreign Languages
基金
电子科技大学青年科学基金的资助
项目编号:Y02001011301005.
