A Multilevel Multidimensional Finite Mixture Item Response Model to Cluster Respondents and Countries
A Multilevel Multidimensional Finite Mixture Item Response Model to Cluster Respondents and Countries
复制标题
对聚类受访者和国家的多级多维有限混合项目响应模型
DOI：
10.1027/1015-5759/a000631
复制
发表时间：
2021
影响因子：
2.5
通讯作者：
Kupeli Nuriye
中科院分区：
文献类型：
--
作者：
Kanovsk? Martin;Halamov? J?lia;Zuroff David C.;Troop Nicholas A.;Gilbert Paul;Shahar Ben;Petrocchi Nicola;Hermanto Nicola;Krieger Tobias;Kirby James N.;Asano Kenichi;Matos Marcela;Yu FuYa;Basran Jaskaran;Kupeli Nuriye
本研究的目的是测试自我批评/攻击和自我安慰量表 (FSCRS) 的多级多维有限混合项目反应模型，对来自 12 个国家的 13 个样本 (N = 7,714) 的受访者和国家进行聚类。实际目标是了解个人（即，要使用多少个截止值）和国家（即，它们之间的相似性和差异性的程度）有多少个离散类别。我们采用了多级多维有限混合方法，该方法基于多维潜在类项目响应理论（IRT）模型的扩展类。个人和国家被划分为具有不同程度的自我批评和自我保证的离散潜在类别，同时考虑到构造的多维结构。该方法适用于使用 FSCRS 的三维测量来分析不同级别（个人和国家）和不同维度的观察到的特征和潜在特征之间的关系。结果表明，受访者的分数取决于未观察到的（潜在类别）个人和国家成员身份、工具的多维结构，并证明了在个人和国家的比较心理评估中使用多级多维有限混合项目反应模型的合理性。 FSCRS 的潜在类别分析表明，个体参与者和国家可以分为离散的类别。根据之前的研究结果，FSCRS 在心理测量上是稳健的，我们可以建议使用 FSCRS 来衡量自我批评。
The aim of this study was to test the multilevel multidimensional finite mixture item response model of the Forms of Self-Criticising/Attacking and Self-Reassuring Scale (FSCRS) to cluster respondents and countries from 13 samples (N= 7,714) and from 12 countries. The practical goal was to learn how many discrete classes there are on the level of individuals (ie, how many cut-offs are to be used) and countries (ie, the magnitude of similarities and dissimilarities among them). We employed the multilevel multidimensional finite mixture approach which is based on an extended class of multidimensional latent class Item Response Theory (IRT) models. Individuals and countries are partitioned into discrete latent classes with different levels of self-criticism and self-reassurance, taking into account at the same time the multidimensional structure of the construct. This approach was applied to the analysis of the relationships between observed characteristics and latent trait at different levels (individuals and countries), and across different dimensions using the three-dimensional measure of the FSCRS. Results showed that respondents’ scores were dependent on unobserved (latent class) individual and country membership, the multidimensional structure of the instrument, and justified the use of a multilevel multidimensional finite mixture item response model in the comparative psychological assessment of individuals and countries. Latent class analysis of the FSCRS showed that individual participants and countries could be divided into discrete classes. Along with the previous findings that the FSCRS is psychometrically robust we can recommend using the FSCRS for measuring self-criticism.