{"title":"具有潜在项目效应变量的等级反应概率多状态IRT模型","authors":"Franz L. Classe, R. Steyer","doi":"10.1027/1015-5759/a000751","DOIUrl":null,"url":null,"abstract":"Abstract. A probit multistate Item Response Theory (IRT) model for ordinal response variables is introduced. It comprises a reference latent state variable for each occasion of measurement and a latent item effect variable for each item except for one reference item. The latent item effect variable is defined as the difference between the latent state variable pertaining to the non-reference item and the latent state variable pertaining to the reference item. They are assumed to be identical for all occasions of measurement. The new model is applied to a real data example. Including item effect variables improve model fit considerably. Hence, the items are not strictly unidimensional within each occasion of measurement.","PeriodicalId":48018,"journal":{"name":"European Journal of Psychological Assessment","volume":null,"pages":null},"PeriodicalIF":3.2000,"publicationDate":"2023-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A Probit Multistate IRT Model With Latent Item Effect Variables for Graded Responses\",\"authors\":\"Franz L. Classe, R. Steyer\",\"doi\":\"10.1027/1015-5759/a000751\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Abstract. A probit multistate Item Response Theory (IRT) model for ordinal response variables is introduced. It comprises a reference latent state variable for each occasion of measurement and a latent item effect variable for each item except for one reference item. The latent item effect variable is defined as the difference between the latent state variable pertaining to the non-reference item and the latent state variable pertaining to the reference item. They are assumed to be identical for all occasions of measurement. The new model is applied to a real data example. Including item effect variables improve model fit considerably. Hence, the items are not strictly unidimensional within each occasion of measurement.\",\"PeriodicalId\":48018,\"journal\":{\"name\":\"European Journal of Psychological Assessment\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":3.2000,\"publicationDate\":\"2023-01-31\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"European Journal of Psychological Assessment\",\"FirstCategoryId\":\"102\",\"ListUrlMain\":\"https://doi.org/10.1027/1015-5759/a000751\",\"RegionNum\":3,\"RegionCategory\":\"心理学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"PSYCHOLOGY, APPLIED\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"European Journal of Psychological Assessment","FirstCategoryId":"102","ListUrlMain":"https://doi.org/10.1027/1015-5759/a000751","RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"PSYCHOLOGY, APPLIED","Score":null,"Total":0}
引用次数: 0
摘要
摘要介绍了一种概率多状态项目反应理论(probit multi - state Item Response Theory, IRT)模型。它包括用于每个测量场合的参考潜在状态变量和用于除一个参考项目外的每个项目的潜在项目效应变量。潜在项目效应变量定义为与非参考项目相关的潜在状态变量与与参考项目相关的潜在状态变量之间的差异。假定它们在所有测量场合都是相同的。将该模型应用于一个实际的数据实例。纳入项目效应变量可显著改善模型拟合。因此,在每次测量中,这些项目并不是严格的单维的。
A Probit Multistate IRT Model With Latent Item Effect Variables for Graded Responses
Abstract. A probit multistate Item Response Theory (IRT) model for ordinal response variables is introduced. It comprises a reference latent state variable for each occasion of measurement and a latent item effect variable for each item except for one reference item. The latent item effect variable is defined as the difference between the latent state variable pertaining to the non-reference item and the latent state variable pertaining to the reference item. They are assumed to be identical for all occasions of measurement. The new model is applied to a real data example. Including item effect variables improve model fit considerably. Hence, the items are not strictly unidimensional within each occasion of measurement.
期刊介绍:
The main purpose of the EJPA is to present important articles which provide seminal information on both theoretical and applied developments in this field. Articles reporting the construction of new measures or an advancement of an existing measure are given priority. The journal is directed to practitioners as well as to academicians: The conviction of its editors is that the discipline of psychological assessment should, necessarily and firmly, be attached to the roots of psychological science, while going deeply into all the consequences of its applied, practice-oriented development.