Anselin Local Moran’s I (Evans, Love & Thurston, 2015; Al-Ahmadi, Alahmadi & Al-Zahrani, 2019; Vilinová, 2020) was used to determine spatial clusters of respiratory ERV incidence and ozone exposures for each quarter during the period 2009–2016. Cluster and outlier classification analysis aids in the discovery of spatial patterns of distinct and similar observations (Evans, Love & Thurston, 2015). The Local Moran’s I statistic evaluates the attribute value (ozone and respiratory ER visits) for each spatial unit on the map (zip code) with the values of that units’ neighbors. The level of influence (spatial weight) assigned to neighbors are a decreasing function of distance, i.e., closest neighbors are weighted more than those further away (Inverse Distance Weighted conceptualization) (Anselin, 1995). Spatial units are categorized into one of four categories: High-High (HH), High-Low (HL), Low-High (LH), and Low-Low (LL). These categories specify the relationship between every zip code on the map and its neighbors. HL and LH categories indicate spatial outliers where the observed spatial unit has a value that is different from its neighbors. HH and LL categories indicate spatial clusters where the observed spatial unit has a value like its neighbors. The output from ArcMap provides p-values for each spatial unit, thus allowing us to distinguish outliers and clusters that are statistically significant (95% confidence level).

注意:以上内容是从某篇研究文章中自动提取的,可能无法正确显示。



Q&A
请登录并在线提交您的问题
您的问题将发布在Bio-101网站上。我们会将您的问题发送给本研究方案的作者和具有相关研究经验的Bio-protocol成员。我们将通过您的Bio-protocol帐户绑定邮箱进行消息通知。