Development of the Neighbourhood Green Space Tool (NGST)

Use of urban green space can confer a range of health benefits, but is thought to be quality dependent. Existing methods of quality assessment might not be appropriate for 'neighbourhood' green space as there are functional differences between these small sites that tend to serve only local residents and larger sites that people travel to visit. Our aim was to develop a simple tool to characterise quality of neighbourhood green space. The study setting was Stoke-on-Trent, UK. Using an existing measure as a template, the tool was developed through: (i) focus groups (n = 35) and piloting; (ii) surveys (n = 635) to finalise items and composite domains, and to determine domain weights that reflected their relative importance; and (iii) testing in 28 sites (mean area 3.0 +/- 2.0 ha) for feasibility and inter-rater reliability using intraclass correlation coefficients (ICC). The 36-item Neighbourhood Green Space Tool (NGST) comprised an overall Usage domain and five domains with respective weights: Accessibility 18.0%; Recreational facilities 16.0%; Amenities 22.0%; Natural features 20.0%; and Incivilities 24.0%. Average time for site assessment indicated high feasibility (11.1 +/- 3.8 min). Reliability was moderate-good for individual domains (ICC = .575-.948, all p < .001) and overall green space quality scores (ICC = .727, p < .001). The NGST demonstrated good feasibility and moderate-good reliability for quality assessment of neighbourhood urban green space. The survey component should be repeated in children and young people to explore differences in preference and subsequent domain weights. Quality should ultimately be validated against level and nature of use. (C) 2012 Elsevier B.V. All rights reserved.