摘要
目的:将基于项目反应理论的计算机自适应测验方式应用于艾森克人格问卷中,分析其可行性和有效性。方法:采用探索性因素分析对量表进行单维性检验,比较项目反应理论模型的优劣,对项目参数和能力参数进行估计;再基于真实的纸笔测验数据模拟计算机自适应测验,分析其测验效率、测验误差,最后将测验分数与纸笔分数进行比较。结果:因子分析结果四个量表大体上能够满足单维性,2PLM优于1PLM;计算机自适应测验结果所有量表都节省了50%以上的项目,所有量表的测验误差均值都低于预设误差值,与纸笔测验分数的相关系数都达到或接近高相关,而校正相关系数除L量表外都在0.9以上。结论:艾森克人格问卷对计算机自适应测验方式是适合的。
Objective: Application of computerized adaptive testing(CAT) based on item response theory to Eysenck Personality Questionnaire(EPQ) was administrated to examine its feasibility and validity.Methods: Exploring factor analysis was used to assess unidimensionality;and appropriate model was chosen to estimate item parameters and person parameters;and then a real data simulation of CAT was used to analyze efficiency of testing and measurement precision and the result of CAT was compared with paper-and-pencil tests.Results: The four scales roughly met unidimensionality,and 2PLM was superior to 1PLM.In the results of CAT,all scales saved 50% items,and measurement errors were below predicted values,the correlations between scores of CAT and conventional paper-and-pencil tests were nearly or above 0.7,and disattenuated correlation coefficients were above 0.9 except for L scale.Conclusion: Application of IRT-based CAT to EPQ is appropriate.
出处
《中国临床心理学杂志》
CSSCI
CSCD
2011年第3期306-308,共3页
Chinese Journal of Clinical Psychology
关键词
计算机自适应测验
艾森克人格问卷
Computerized adaptive testing
Eysenck Personality Questionnaire