目的：同结构测验的编制与验证方法的研究不仅可促进动态评估的推广和应用，同时也将为其它测验的编制提供参考。方法：基于标准图形推理测验项目的认知模型研究结果，编制了三套同结构标准图形推理测验。每套测验均包含21个项目，四种项目结构。从江西省北部抽取230名儿童参加这三套测验的测试。然后把项目结构类型作为潜因子，把项目作为观察变量，采用验证性因素分析方法(CFA)对这三套测验进行了结构验证。结果: 这三套测验在结构上是相同的。结论：项目认知模型不仅是编制同结构测验的认知心理学基础，也是提供基于项目认知加工过程编制测验的新思想和新方法。

[1] Embretson SE.A multidimensional latent trait model formeasuring learning and change.Psychometrika,1991,56(3):495-515;
[2] Embretson SE.Toward development of a psychometric approach.In Carol SL.Dynamic Assessment:An iteractionalapproach to evaluating learning potential.The GuilfordPress.NewYork and London,1987.141-161;
[3] Cronbach LJ,Furby L.How should we measure change orshould we? Psychological Bulletin,1970,74:68-80;
[4] Embretson SE.Multidimensional measurement from dynamictests:Abstract reasoning under stress.Multivariate BehavioralResearch,2000,35(4):505-542;
[5] Embretson SE.A cognitive design system approach to generatingvalid tests:Application to abstract reasoning.PsychologicalMethods,1998,3(3):380-396;
[6] Embretson SE.Generating items during testing:Psychometricissues and models.Psychometrika,1999,64(4):407-433;
[7] Embretson SE,Yang XD.Automatic item generation andcognitive psychology.Handbook of statistics,2007,26:47-76;[6] ;
[8] Carpenter PA,Just MA,Shell P.What one intelligence testmeasures:A theoretical account of the processing in theraven progressive matrices test.Psychological Review,1990,4(3):404-443;
[9] Embretson SE.Cognitive test design systems:Application toabstract reasoning.Paper presented at the 1995 annualmeeting of the American Education Research Association.April,San Francisco,CA,1995
[10] 戴海琦,刘声涛.瑞文推理测验项目认知难度因素分析及LLTM 拟合验证.心理与行为研究,２００４,2(2):411-414
[11] 周骏,戴海琦,徐淑媛,康春花.用于项目生成的认知模型的构建与比较.心理学探新,２０１０,30(3A):56-62
[12] 周骏,戴海琦,徐淑媛.矩阵完成问题的项目生成研究.心理与行为研究,２０１０,8(3B):166-171
[13] 候杰泰,温忠麟,成子娟.结构方程模型及其应用.北京:教育科学出版社,２００４.25-51,90,154-165
[14] Hu L,Bentler PM.Fit indeices in covariance structuremodeling:Sensitivity to underparameterized model misspecification.Psychological Methods,１９９８,3:424-453