Download Now Free registration required
Using GPS devices and digital cameras, the authors surveyed outdoor advertisements in Austin, Los Angeles and Philadelphia. GIS and hot spot analysis revealed that unhealthy ads were clustered around child-serving institutions in Los Angeles and Philadelphia but not in Austin. Multivariate generalized least square (GLS) regression models showed that percent black was a significant positive predictor of clustering in Philadelphia and percent white was a marginally significant negative predictor of clustering in Los Angeles after controlling for several land use variables.
- Format: PDF
- Size: 257.2 KB