Queen Mary University of London.
Urban Stud. 2011;48(10):2101-22. doi: 10.1177/0042098010380961.
Analysis of geographical patterns of suicide and psychiatric morbidity has demonstrated the impact of latent ecological variables (such as deprivation, rurality). Such latent variables may be derived by conventional multivariate techniques from sets of observed indices (for example, by principal components), by composite variable methods or by methods which explicitly consider the spatial framework of areas and, in particular, the spatial clustering of latent risks and outcomes. This article considers a latent random variable approach to explaining geographical contrasts in suicide in the US; and it develops a spatial structural equation model incorporating deprivation, social fragmentation and rurality. The approach allows for such latent spatial constructs to be correlated both within and between areas. Potential effects of area ethnic mix are also included. The model is applied to male and female suicide deaths over 2002–06 in 3142 US counties.
对自杀和精神疾病发病率的地理模式的分析表明了潜在生态变量（如贫困、农村）的影响。这些潜在变量可以通过常规多元技术从一组观察指标中得出（例如，通过主成分分析），也可以通过综合变量方法或明确考虑区域空间框架的方法得出，特别是要考虑潜在风险和结果的空间聚类。本文考虑了一种潜在随机变量方法来解释美国自杀的地理差异；并开发了一个包含贫困、社会分裂和农村的空间结构方程模型。该方法允许在区域内和区域之间相关联这些潜在的空间结构。还包括区域民族构成的潜在影响。该模型应用于 2002-06 年美国 3142 个县的男性和女性自杀死亡人数。
