The results of investments in transit and redevelopment vary widely across space. To better understand the investment and use connection, this research analyzes the spatial characteristics of modal choice and land-use locally rather than globally. The proportion of people around stations using transit is modeled as a function of environmental variables, spatial proximity to transit, and spatial autocorrelation among those variables by using geographically weighted regression (GWR) on data from St. Louis, Missouri. The analysis generated spatially variant regression coefficients and R-square values that suggest significant spatial variation in the influence of neighborhood factors on modal choice.
Available at: http://works.bepress.com/bradleywlane/9/