Beyond representing your data in a map, GIS can help you to answer questions by using spatial analysis methods. Where have case rates been increasing over the last ten years? What factors are associated with county-level rates of low birth weights? How many people in Texas live more than an hour away from a dental clinic?
Spatial analysis is the process of exploring the locations, attributes, and relationships of features in spatial data to answer questions or better understand phenomena. Some common objectives that we can help you to address through spatial analysis include:
Describe features and distributions
• Aggregate data and create quantitative thematic maps
• Generate spatial mean and standard deviational ellipses to summarize point locations
Example: Choropleth map of population to licensed vocational nurse ratio by county for 2019. Counties in white had no licensed vocational nurses.
Determine how places are related
• Find the nearest facility for a given population
• Identify areas outside of a specific distance or driving time of service provider
Example: Drive-time analysis showing areas within different driving times of the hospital emergency department in Fredericksburg, Texas. Darker areas indicate a short drive.
Identify and quantify patterns
• Determine if there is spatial clustering in the distribution of point locations
• Identify spatial outliers (locations with very high or low values compared to their neighbors)
Example: Hot spot analysis of Social Vulnerability Index at the census tract level in Northcentral Texas
Find places that meet certain criteria
• Perform spatial overlay analysis to select the best location for a new facility or project
• Identify areas where something is most likely to occur
Example: Selection of census tracts with a Social Vulnerability Index > 0.75 that also contain a hospital around Bexar County
Explain or predict an outcome
• Use spatial interpolation to estimate missing values and create a smooth surface
• Identify and characterize different types of emerging hotspots and creating future projections
• Regression-based modelling, including geographically weighted regression
Example: Projected probability of West Nile virus detection based on logistic regression model in Hunt County