Assessing urban and rural neighborhood characteristics using audit and GIS data: derivation and reliability of constructs Public Deposited
Downloadable ContentDownload PDF
- Affiliation: Gillings School of Global Public Health, Department of Epidemiology
- Affiliation: Gillings School of Global Public Health, Department of Biostatistics
Laraia, Barbara A
- Other Affiliation: Department of Medicine, Division of Prevention Sciences, Center for Health and Community, University of California, San Francisco, CA, USA
- Other Affiliation: Center for Health Policy, Duke Global Health Institute, Duke University, Durham, NC, USA
- Affiliation: College of Arts and Sciences, Department of City and Regional Planning
- Affiliation: Gillings School of Global Public Health, Department of Biostatistics, Carolina Population Center
- Abstract Background Measures to assess neighborhood environments are needed to better understand the salient features that may enhance outdoor physical activities, such as walking and bicycling for transport or leisure. The purpose of this study was to derive constructs to describe neighborhoods using both primary (neighborhood audit) and secondary (geographic information systems) data. Methods We collected detailed information on 10,770 road segments using an audit and secondary data. The road segment sample was randomly split into an exploratory (60%) and validation sample (40%) for cross-validation. Using the exploratory sample (n = 6,388), seven a priori constructs were assessed separately (functionality, safety, aesthetics, destinations, incivilities, territorality, social spaces) by urbanicity using multi-group confirmatory factor analysis (CFA). Additionally, new a posteriori constructs were derived using exploratory factor analysis (EFA). For cross-validation (n = 4,382), we tested factor loadings, thresholds, correlated errors, and correlations among a posteriori constructs between the two subsamples. Two-week test-retest reliability of the final constructs using a subsample of road segments (n = 464) was examined using Spearman correlation coefficients. Results CFA indicated the a priori constructs did not hold in this geographic area, with the exception of physical incivilities. Therefore, we used EFA to derive a four-factor solution on the exploratory sample: arterial or thoroughfare, walkable neighborhood, physical incivilities, and decoration. Using CFA on the validation sample, the internal validity for these a posteriori constructs was high (range 0.43 to 0.73) and the fit was acceptable. Spearman correlations indicated the arterial or thoroughfare factor displayed near perfect reliability in both urban and rural segments (r = 0.96). Both the physical incivilities factor and the walkable neighborhood factor had substantial to near perfect reliability in both urban and rural segments (r = 0.77 to 0.78 and r = 0.79 to 0.82, respectively). The decoration factor displayed moderate reliability in urban segments (r = 0.50; 95% CI: 0.38–0.60) and lower reliability in rural segments (r = 0.39; 95% CI: 0.25–0.52). Conclusion The results of our analyses yielded four reliably and objectively measured constructs that will be used to explore associations with physical activity in urban and rural North Carolina. These constructs should be explored in other geographic areas to confirm their usefulness elsewhere.
- Date of publication
- July 20, 2009
- Resource type
- Rights statement
- In Copyright
- Rights holder
- Kelly R Evenson et al.; licensee BioMed Central Ltd.
- Journal title
- International Journal of Behavioral Nutrition and Physical Activity
- Journal volume
- Journal issue
- Page start
- Is the article or chapter peer-reviewed?
- Bibliographic citation
- International Journal of Behavioral Nutrition and Physical Activity. 2009 Jul 20;6(1):44
- Open Access
- BioMed Central Ltd
This work has no parents.
|PIN3 Neighborhood Audit Instrument. This file provides the questions used on the PIN3 Neighborhood Audit instrument.||2019-05-06||Public||
|Descriptive statistics for GIS and PIN3 Neighborhood Audit variables for entire sample (n = 10,770). This file provides descriptive statistics for all of the GIS and neighborhood audit variables used in the analyses.||2019-05-06||Public||