Comparison and diagnostics of various latent variable models in social sciences

来源 :The 24th International Workshop on Matrices and Statistics(第 | 被引量 : 0次 | 上传用户:cmccetehi
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
  Different kinds of questionnaires are usually applied in a field of social sciences. The basic interest of these studies is often to reveal the underlying construct which is measured by different set of specific questions. Usually the construct of measurement instrument is examined using latent variable model e.g. exploratory or confirmatory factor analysis, (multidimensional) item response models (Bock, Gibbons & Muraki, 1988), latent class or latent profile analysis (Goodman, 1974), depending on the measurement level of observed variables and the assumptions of underlying model. Although the latent variable models are well known and usually part of basic methodological curriculum, many still struggle with estimation problems e.g. Haywood cases and non-identifiability. To identify the cause of these problems one must carefully examine the huge number of different result matrixes, which can be rather difficult in most commercial software. At the same time different models can lead to similar results and its often difficult to choose which model is best for specific problem or data. In this presentation Ill show how Survo-R environment (Sund, Vehkalahti & Mustonen, 2014) might help researcher to solve different computational issues, which usually arise when one is trying to analyze real word data leading to a better understanding of underlying model. Survo-R is a powerful editor of an R-programing language (R Core Team, 2013) which is of the most widely used statistical environment.
其他文献
  With a history of more than 3000 years, magic squares still are mysterious in various aspects. We in this paper give a comprehensive review and study on cla
会议
  Our motivation in this talk is the 13th-century Anxi iron-plate doubly-classic 6x6 bordered magic square discussed by Kai-Tai Fang at the 22nd International
会议
  The R2 statistic in fixed-effects regression settings is routinely interpreted as a measure of proportion of variability explained by the model. Because R2
会议
  I will reflect on selected experiences with Simo Puntanen concentrating especially on the scientific adventures we have shared around the globe during the l
会议
  Low response rate characterizes nowadays sample surveys. Furthermore, the resulting response set is biased. Special adjustment methods are needed to reduce
会议
  My recent book Antieigenvalue Analysis , World-Scientific, 2012, presented the theory of antieigenvalues from its inception in 1966 up to 2010, and its appl
会议
  This is what I try to figure out in this talk.
会议
  Infections during pregnancy will increase womens risk of serious consequences. People have started to study the cohorts with safety data for vaccination dur
会议
  Model predictive control is a widely used industrial technique to deal with trajectory tracking problems in many process industry applications, as well as i
  Matrix computations form a core for many of the traditional multivariate statistical methods. Especially the matrix decompositions, such as singular value d
会议