Fig. 3

Estimated mean a of percent of supermarkets relative to total supermarkets, grocery stores and convenience stores by six types of neighborhoods b: Twin Cities Region, 1993–2011. a Multivariable mixed effects regression modeling percent of supermarkets relative to total supermarkets, grocery stores and convenience stores in each neighborhood as a function of neighborhood type in 1993, time elapsed since 1993, interaction between neighborhood type in 1993 and time elapsed, changes in residential population density, median household income, percent of white and percent of single-family housing since 1993, total supermarkets, grocery stores and convenience stores and a random intercept for each neighborhood. b Derived from cluster analysis of block-group level data in 1993: percent of age under 14, aged 15–29, 30–44, 45–64, and aged above 65, percent of education of college or above, percent of white, percent of black, median household income, residential population density, employment population density, mix of land use and percent of single-family housing