当前位置: X-MOL 学术AStA. Adv. Stat. Anal. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Varying levels of anomie in Europe: a multilevel analysis based on multidimensional IRT models
AStA Advances in Statistical Analysis ( IF 1.4 ) Pub Date : 2018-02-10 , DOI: 10.1007/s10182-018-0320-0
Lara Fontanella , Annalina Sarra , Pasquale Valentini , Simone Di Zio , Sara Fontanella

Recent years have seen increased attention paid to monitoring social anomie and its dependency on micro- and macro-factors. In this paper, we endorse the theorisation of social anomie as a complex, multidimensional and multilevel phenomenon. To ensure a rigorous measurement of the varying levels of social anomie in the European countries, the current study relies on a multilevel multidimensional item response theory model which explicitly accounts for the presence of a non-ignorable missing data mechanism. This unified approach makes it possible to specify an analytical model of links between anomie features and their determinants and to explore how the latent traits of interest are influenced by individual-level factors, as well as by country-level indicators. Additionally, to avoid misleading inferential conclusions, the proposed model takes into account the respondent’s omitting behaviour, assuming that the missingness mechanism is driven by a latent propensity to respond. Data used in this study have been collected in the 2010 wave of the European Social Survey. To reduce the computational complexities, a Bayesian specification of the MIRT model is provided and the parameter model estimates are obtained through MCMC algorithms.

中文翻译:

欧洲不同的失范水平:基于多维IRT模型的多层次分析

近年来,人们越来越重视监视社会失范及其对微观和宏观因素的依赖性。在本文中,我们赞同将社会失范理论化为一个复杂的,多维的,多层次的现象。为了确保严格测量欧洲国家社会失范的不同水平,当前的研究依赖于多维多维项目响应理论模型,该模型明确说明了不可忽略的缺失数据机制的存在。这种统一的方法使得可以指定一个异常特征及其决定因素之间联系的分析模型,并探索个人潜在因素和国家层面指标如何影响感兴趣的潜在特征。此外,为避免误导性推断结论,假设缺少机制是由潜在的响应倾向驱动的,建议的模型考虑了受访者的遗漏行为。本研究中使用的数据已在2010年欧洲社会调查浪潮中收集。为了降低计算复杂度,提供了MIRT模型的贝叶斯规范,并通过MCMC算法获得了参数模型估计值。
更新日期:2018-02-10
down
wechat
bug