Title: Regional assessment of environmental equity through GIS-based clustering and non-parametric statistical testing: a case study of Dallas County, Texas, USA

Authors: Minhe Ji, T.S. Sunil

Addresses: Department of Geography, University of North Texas, P.O. Box 305279, Denton, TX 76203-5279, USA. ' Department of Sociology, University of Texas at San Antonio, San Antonio, TX 78249, USA

Abstract: Previous GIS-based methods for assessing regional environmental equity issues tended to simplify the definition of impact area for a hazardous site by either directly adopting the pre-existing census unit or constructing a buffer with an arbitrary radius. Subsequent spatial analyses based on such a simplification support rather limited statistical analysis. The present paper reviews this methodological issue and proposes a more robust method combining statistical clustering, spatial overlay analysis, and non-parametric statistical testing for regional equity assessment. In this method, impact areas for individual population groups are derived through statistical clustering based on the internal data structure of multiple demographic variables, and facilities are tallied within each cluster zone to support a valid chi-square test. A case study on Dallas County, Texas, USA, demonstrated the advantages of this method, including data self-organisation for case-specific analysis, avoidance of arbitrariness of impact area definition, ability to provide a comprehensive evaluation based on multiple demographic characteristics, and statistical comparability with highly skewed data distribution.

Keywords: environmental equity assessment; GIS; spatial analysis; statistical clustering; spatial overlay analysis; non-parametric statistical testing; regional equity assessment; geographical information systems; hazard assessment; risk assessment.

DOI: 10.1504/IJRAM.2005.006610

International Journal of Risk Assessment and Management, 2005 Vol.5 No.1, pp.36 - 49

Published online: 28 Mar 2005 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article