摘要
主观考试采用评分员进行主观评分，由于评分一致性不高，缺乏信度，测量学界一直在努力探索提高主观评分信度的办法。本文用Longford方法对参加HSK[高等]作文考试评分的异常评分员作了一次实证检验。结果证明，该方法对检验大规模标准化主观考试评分员差异确实有效。
Experts in testing field have been making efforts to promote the reliability of subjective rating since rating errors exist among raters, which leads to lower rating reliability. This paper conducted an empirical experiment on judging the disagreements among subjective raters by using the method introduced by Longford, an expert on subjective rating, in 1995.
出处
《中国考试》
2010年第1期22-27,共6页
journal of China Examinations
关键词
主观评分
评分员
评分差异
主观评分信度
Subjective Rating
Rater
Rating Errors
Subjective Rating Reliability
