China
Africa
Explore chapters and articles related to this topic
Data Exploration
Published in Terry A. Slocum, Robert B. McMaster, Fritz C. Kessler, Hugh H. Howard, Thematic Cartography and Geovisualization, 2022
Terry A. Slocum, Robert B. McMaster, Fritz C. Kessler, Hugh H. Howard
GeoDa is intended for the exploratory analysis of spatial data, with an emphasis on examining the statistical significance of that data. The initial version was developed by Luc Anselin and his colleagues at the University of Illinois, Urbana-Champaign. The most recent version is now available through The Center for Spatial Data Science at the University of Chicago (https://spatial.uchicago.edu/geoda).
Spatial variability of trace elements with Moran’s I Analysis for shallow groundwater quality in the Lower Katari Basin, Bolivian Altiplano
Published in Yong-Guan Zhu, Huaming Guo, Prosun Bhattacharya, Jochen Bundschuh, Arslan Ahmad, Ravi Naidu, Environmental Arsenic in a Changing World, 2019
I. Quino, O. Ramos, M. Ormachea, J. Quintanilla, P. Bhattacharya
The Moran’s I statistic was used with LISA (Local Indicators of Spatial Association) method to know the spatial autocorrelation (SA) of each element. The global spatial dependence analysis (Global Moran’s I statistiscal test), the local spatial dependence (BiLISA Cluster Map) and the significant spatial test (BiLISA Significance Map) were made in the GeoDa 1.12.01 software.
Spatial analysis and evaluation of road traffic safety performance indexes across the provinces of Turkey from 2015 to 2019
Published in International Journal of Injury Control and Safety Promotion, 2021
Morteza Ahmadpur, Ilgin Gokasar
GeoDa software was employed for conducting spatial analysis and clustering. Regression and correlation analysis were conducted by using SPSS software. Figures and choropleth maps illustrating the geographic distribution of studied indexes and results of clustering were produced by using GeoDa software.
Hot spot analysis and evaluation of influencing factors on regional road crash safety and severity indices: insights from Iran
Published in International Journal of Injury Control and Safety Promotion, 2023
Moran’s I, an index between −1 and +1, measures spatial autocorrelation’s strength and helps detect regions’ spatial patterns in terms of studied variables. A positive and high Moran’s I value suggests that the distribution of the provinces is cluster-like. If Moran’s I is negative and approaches −1, it implies that the studied provinces have a random distribution in terms of the studied index. Variables with negative Moran’s I were removed from the correlation tables. The local Moran statistic is a local indicator of spatial association (Anselin, 2010) and is an index to identify significant local clusters and local spatial outliers. By considering the calculated local Moran indices, provinces were categorised as hot spots (high-high [HH] significant spatial clusters or cluster cores), cold spots (low-low [LL] significant spatial clusters), outliers (high-low [HL] or low-high [LH]), and non-significant regions. For example, a significant HH cluster means that provinces with high and similar indices surround this HH province and form a hot spot. Spatial weights are critical in constructing spatial autocorrelation statistics such as Moran’s I. These weights represent the neighbour structure between observations. In this study, spatial contiguity weights were adopted, and contiguity means that two provinces share a common border of nonzero length. In the evaluation process of the local Moran indices, a p-value less than 0.01 was adopted and hot and cold spots were significant at this p-value for 999 permutations in GeoDa software, except for the TR index (p-value < 0.05). Since the TR is an important index in road safety studies, in calculating the local Moran index of this RSI, instead of 0.01, a p-value less than 0.05 was selected. In calculating the local Moran index, selecting the p-value is important, and the conventional choice of 0.05 is prone to cause false positives (Anselin, 2005). Therefore, the p-value was set to 0.01 to decrease the number of false positives in this study. Spatial analysis, clustering, and mapping were conducted by using GeoDa software.