Regression Toward the Mean in Neighborhood Effects Research
A Geographic Perspective
Yinhua Tao (Southeast University)
A Petrović (TU Delft - Urban Studies)
Mei Po Kwan (Chinese University of Hong Kong)
M van Ham (TU Delft - Urban Studies)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Neighborhood effects research focuses on the residential neighborhood, assuming it as the main spatial context relevant to individual outcomes. Individuals, however, are mobile and visit various spatial contexts other than the residential neighborhoods. This article conceptualizes contextual exposures to socioenvironmental factors in daily activity spaces and their relationship with residential exposures. By introducing regression toward the mean, we argue that mobility-based contextual exposures are, on average, less extreme than residential exposures. Previous neighborhood effects studies therefore tend to underestimate actual spatial contextual effects when they misrepresent residential neighborhood effects as the total contextual effects. Despite improved measurement accuracy with the transition from residence- to mobility-based exposures, we suggest the complexities remaining in the estimation of spatial contextual effects from a geographic perspective. These complexities include a possibly limited extent of neighborhood effects regression across neighborhoods and asymmetrical dispersion of between-individual contextual exposures within each neighborhood.
Files
File under embargo until 03-12-2025