Minimizing aggregation errors when measuring potential access to services for social groups at the city scale. (October 2021)

• Record Type: Journal Article
• Authors: Wu, Chengcheng; Powe, Neil A; Copeland, Alison
• Is Part Of: Environment & planning. Volume 48:Number 8(2021)
• Page Start: 2206
• Page End: 2220
• Publication Date: 2021-10
• Subjects: Ecological fallacy and Modifiable Areal Unit Problem -- aggregation errors -- place access and population access -- potential accessibility and potential access measurement -- services -- social groups; City planning -- Periodicals; Urban ecology (Sociology) -- Periodicals; 307.11605
• DOI: 10.1177/2399808320970201
• Languages: English
• ISSNs: 2399-8083

Abstract

This research explores how to minimize aggregation errors when measuring potential access to services for social groups at the city scale. It develops a cadastral and address-based population weighting technique, the Household Space Weighting, to reduce aggregation errors caused by using population weighted centroids when applying the Have Their Centre In criterion (the Population Weighted Centroid technique). The Household Space Weighting technique is formally tested in a case study of General Practitioner practices in Newcastle upon Tyne, UK. The findings suggest that the Population Weighted Centroid technique produces inaccurate population estimates for 267 out of 910 output areas (29%) in the city. When applying the two techniques to measure access for social groups at the city scale, the absolute difference in the percentage of each social group with potential accessibility is 9–10% and the relative difference in the percentage of each social group with potential access is 18–20%, taking into account the overlay of service areas at the city scale. This suggests that if service planners or policy makers want to measure potential accessibility or potential access of social groups to services for cities, it would be useful to apply a more accurate technique, or at least be aware of the implications of applying the Population Weighted Centroid technique.