Department of Allied Health Sciences, University of North Carolina, CB #7122, Bondurant Hall, Suite 2050, Chapel Hill, NC 27599-7122, USA.
Health Place. 2011 Sep;17(5):1113-21. doi: 10.1016/j.healthplace.2011.05.011. Epub 2011 Jun 7.
The literature on neighborhoods and health highlights the difficulty of operationalizing "neighborhood" in a conceptually and empirically valid manner. Most studies, however, continue to define neighborhoods using less theoretically relevant boundaries, risking erroneous inferences from poor measurement. We review an innovative methodology to address this problem, called the socio-spatial neighborhood estimation method (SNEM). To estimate neighborhood boundaries, researchers used a theoretically informed combination of qualitative GIS and on-the-ground observations in Texas City, Texas. Using data from a large sample, we assessed the SNEM-generated neighborhood units by comparing intra-class correlation coefficients (ICCs) and multi-level model parameter estimates of SNEM-based measures against those for census block groups and regular grid cells. ICCs and criterion-related validity evidence using SF-36 outcome measures indicate that the SNEM approach to operationalization could improve inferences based on neighborhoods and health research.
关于邻里关系和健康的文献强调了以概念上和经验上有效的方式操作“邻里”的困难。然而，大多数研究仍然使用理论上相关性较低的边界来定义邻里关系，从而有可能因测量不佳而得出错误的推论。我们回顾了一种解决这个问题的创新方法，称为社会空间邻里估计方法（SNEM）。为了估计邻里关系的边界，研究人员在德克萨斯州的得克萨斯城使用了一种基于理论的定性 GIS 和实地观察的组合。使用来自大样本的数据，我们通过比较基于 SNEM 的措施的组内相关系数（ICC）和多层次模型参数估计值与人口普查街区组和常规网格单元的对应值，来评估 SNEM 生成的邻里单元。基于 SF-36 结果测量的 ICC 和标准相关有效性证据表明，基于 SNEM 的操作方法可以改善基于邻里关系和健康研究的推论。
