兼顾两种匹配的考试抄袭统计量研究评述  

Review on the Research of Answer-copying Statistics based on the Matching Incorrect and Correct Answers

作　　者：甘媛源[1] 田金亭[1] 余嘉元[1] 

机构地区：[1]南京师范大学心理学系,南京210097

出　　处：《心理学探新》2012年第1期86-90,共5页Psychological Exploration

基　　金：国家社会科学基金"十一五"规划课题(BBA080050)

摘　　要：考试抄袭统计量是专门用于识别抄袭的统计指标。它们是以被怀疑抄袭者和抄袭来源匹配反应的数目为基础建立的,按其考虑的匹配反应信息可分为两类:一类只考虑被怀疑抄袭者和抄袭来源间的错误答案匹配;另一类同时考虑两者错误答案匹配和正确答案匹配。由于第一类抄袭统计量对那些只从抄袭来源处抄袭正确答案的抄袭者检测不敏感,所以存在缺陷;第二类抄袭统计量兼顾两种匹配,考虑的信息更全面。s2、g2、ω是具有代表性的第二类抄袭统计量,它们在测量理论基础、被怀疑抄袭者和抄袭来源的异常相似反应模式确定方法及考试抄袭检测效能等方面存在差异,在抄袭识别应用中应合理选择。Answer - copying statistics are especially presented to detect answer copying, and the likelihood of two examinees suspicion given the same answer pattern is calculated. According to the same answer pattern considered of two examinees, they could be classified two：the copying indices based solely on the matching incorrect answers and that based on both the matching correct answers and the matching incorrect answers. The first one is not sensitive to the copier who copies only the correct answers of the source, and the second one like s2, g2, and ω overcomes the limitation, which are different in three aspects：the measurement theory, the methods on determining the likelihood of two examinees given the same answer pattern and the detection power. Therefore, it is necessary to choose the appropri- ate copying index in practice.

关 键 词：抄袭统计量 K指数 s2指数 g2指数 ω指数 

分 类 号：B841.2[哲学宗教—基础心理学]