Regression and Spatial Analysis of 911 Calls Using ArcGIS Pro

Tools Used: ArcGIS Pro
Skills Applied: Ordinary Least Squares (OLS), Explanatory Regression, Geographically Weighted Regression (GWR), Spatial Autocorrelation
Dataset: 911 Call Incident Data
Course: GEOG 6700 – Quantitative Methods and Spatial Analysis

Project Overview:
In this project, I analyzed the spatial distribution of 911 calls to understand the relationship between emergency incidents and socio-economic variables. Using Ordinary Least Squares (OLS), I first tested individual and multiple explanatory variables including population, job count, low education percentage, and distance to urban center. I then applied Explanatory Regression to evaluate multiple variable combinations and select the best-fitting model based on Adjusted R², AIC, and residual statistics.

To account for spatial variation, I used Geographically Weighted Regression (GWR), which revealed that the influence of population and other variables varied significantly across space. Compared to the global model, GWR improved the model's performance with a lower AIC and a higher Adjusted R² of 0.86, indicating strong explanatory and predictive power. The final spatial prediction map identified future high-risk areas for 911 calls, aiding in proactive emergency response planning.