Theory suggests that perception of neighborhood quality is influenced by crime, blight, and other outdoor characteristics, by respondent perception of the quality of their housing, and by attributes of the local jurisdiction. It also suggests that these geographical attributes should be confounded by respondent demographic characteristics. Using 48,000 sample responses collected by the American Housing Survey in 2002, ordinary least squares, factor analysis, and hierarchical linear modeling, we connected these strands of theory and found strong associations between neighborhood quality, detrimental conditions, housing quality, socioeconomic status, and age. We also measured a small anti-big city effect, and a few weaker relationships with growing/healthy and demographic homogeneity effect. The application of complex statistical tools, in short, enabled us to obtain insight about how people build mental models of their neighborhoods.