摘要
考试作弊甄别是指考试后基于考生作答信息识别异常作答考生的一种防作弊手段。目前针对考生个体作弊的甄别技术主要有答案抄袭检测法和个人拟合检测法。系统梳理2类方法的原理、常见指标、优点、局限性以及组合使用的优势,为考试机构开展作弊甄别工作提供技术参考。
In order to provide technical reference for the examination institutions to carry out the cheating detection work,this paper systematically summarizes the principles,common indice,advantages,limitations and the combinations of different methods at individual level.At present,common detection methods include answer copy detection and person-fit detection.Answer copy detection method is to judge the suspicion of cheating based on the consistency of two examinees,the higher the abnormal consistency of the answers,the greater suspicion of cheating.Common answers copy detection indices based on the above principles include K,K1,K2,S1,S2,g1,g2 andωetc.The person-fit detection method measures the homogeneity of the examinee’s responses.If the examinee shows great inconsistency in his/her abilities in different parts of one test,then the homogeneity of the examinee’s responses is poor.Common person-fit index includes KL divergence and lz index etc.On the basis of many existing cheating detection indices,some researchers put forward a method of integrating different indices so as to maximize the detection effect under different conditions.For example,Belov and Armstrong combine KL divergence and K index to detect examinee’s person-fit and answer copy together,and use a variety of information to improve the detection accuracy and index applicability.Against the background of diversified and high-tech cheating methods,on-site prevention and control are becoming more and more difficult.The method of detecting abnormal examinee through cheating detection statistical techniques will play an increasingly important role.
作者
胡佳琪
黄美薇
骆方
HU Jiaqi;HUANG Meiwei;LUO Fang(Beijing Normal University,Beijing 100875,China)
出处
《中国考试》
CSSCI
2020年第11期32-36,共5页
journal of China Examinations
基金
中国基础教育质量监测协同创新中心2018年度研究生自主课题“作弊的甄别技术现状研究”(BJZK-2018A2-18001)。
关键词
考试作弊甄别
个体作弊
答案抄袭检测
个人拟合检测
考试安全
detection of cheating in examination
individual cheating
answer copy detection
person-fit detection
examination security
作者简介
胡佳琪(1996-),女,北京师范大学心理学部,在读硕士生;黄美薇(1990-),女,北京师范大学中国基础教育质量监测协同创新中心,研究助理;骆方(1979-),女,北京师范大学心理学部,教授。
