Validation of Walk Scores and Transit Scores for estimating neighborhood walkability and transit availability

A small-area analysis

Dustin Duncan, Jared Aldstadt, John Whalen, Steven J. Melly

Research output: Contribution to journalArticle

Abstract

We investigated the validity of Walk Scores and Transit Scores from the Walk Score website using several objective geographic information systems (GIS) measures of neighborhood walkabiltiy and transit availability based on 400- and 800-m street network buffers. Address data come from the 2008 Boston Youth Survey Geospatial Dataset, a school-based sample of public high school students in Boston, MA with complete residential address information (n = 1,292). GIS data were used to create multiple objective measures of neighborhood walkability and transit availability. We also obtained Walk Scores and Transit Scores. We calculated Spearman correlations of Walk Scores and Transit Scores with the GIS neighborhood walkability/transit availability measures as well as Spearman correlations accounting for spatial autocorrelation. Several significant correlations were observed between Walk Score and 400-m buffer GIS measures of neighborhood walkability; all significant correlations were found for the 800-m buffer. All correlations between Transit Scores and GIS measures of neighborhood transit availability were also significant (all p < 0. 0001). However, the magnitude of correlations varied by the GIS measure and neighborhood definition. Relative to the 400-m buffer, correlations for the 800-m buffer were higher. This study suggests that Walk Score is a good, convenient tool to measure certain aspects of neighborhood walkability and transit availability (such as density of retail destinations, density of recreational open space, intersection density, residential density and density of subway stops). However, Walk Score works best at larger spatial scales.

Original languageEnglish (US)
Pages (from-to)407-416
Number of pages10
JournalGeoJournal
Volume78
Issue number2
DOIs
StatePublished - 2013

Fingerprint

information system
open space
analysis
autocorrelation
residential density
geographic information system
student
school
website

Keywords

  • Neighborhood walkability
  • Small-area analysis
  • Transit availability
  • Transit score
  • Validity
  • Walk Score

ASJC Scopus subject areas

  • Geography, Planning and Development

Cite this

Validation of Walk Scores and Transit Scores for estimating neighborhood walkability and transit availability : A small-area analysis. / Duncan, Dustin; Aldstadt, Jared; Whalen, John; Melly, Steven J.

In: GeoJournal, Vol. 78, No. 2, 2013, p. 407-416.

Research output: Contribution to journalArticle

@article{4eafae0820334c6eb1a4d833c5aa721c,
title = "Validation of Walk Scores and Transit Scores for estimating neighborhood walkability and transit availability: A small-area analysis",
abstract = "We investigated the validity of Walk Scores and Transit Scores from the Walk Score website using several objective geographic information systems (GIS) measures of neighborhood walkabiltiy and transit availability based on 400- and 800-m street network buffers. Address data come from the 2008 Boston Youth Survey Geospatial Dataset, a school-based sample of public high school students in Boston, MA with complete residential address information (n = 1,292). GIS data were used to create multiple objective measures of neighborhood walkability and transit availability. We also obtained Walk Scores and Transit Scores. We calculated Spearman correlations of Walk Scores and Transit Scores with the GIS neighborhood walkability/transit availability measures as well as Spearman correlations accounting for spatial autocorrelation. Several significant correlations were observed between Walk Score and 400-m buffer GIS measures of neighborhood walkability; all significant correlations were found for the 800-m buffer. All correlations between Transit Scores and GIS measures of neighborhood transit availability were also significant (all p < 0. 0001). However, the magnitude of correlations varied by the GIS measure and neighborhood definition. Relative to the 400-m buffer, correlations for the 800-m buffer were higher. This study suggests that Walk Score is a good, convenient tool to measure certain aspects of neighborhood walkability and transit availability (such as density of retail destinations, density of recreational open space, intersection density, residential density and density of subway stops). However, Walk Score works best at larger spatial scales.",
keywords = "Neighborhood walkability, Small-area analysis, Transit availability, Transit score, Validity, Walk Score",
author = "Dustin Duncan and Jared Aldstadt and John Whalen and Melly, {Steven J.}",
year = "2013",
doi = "10.1007/s10708-011-9444-4",
language = "English (US)",
volume = "78",
pages = "407--416",
journal = "Geo Journal",
issn = "0343-2521",
publisher = "Springer Netherlands",
number = "2",

}

TY - JOUR

T1 - Validation of Walk Scores and Transit Scores for estimating neighborhood walkability and transit availability

T2 - A small-area analysis

AU - Duncan, Dustin

AU - Aldstadt, Jared

AU - Whalen, John

AU - Melly, Steven J.

PY - 2013

Y1 - 2013

N2 - We investigated the validity of Walk Scores and Transit Scores from the Walk Score website using several objective geographic information systems (GIS) measures of neighborhood walkabiltiy and transit availability based on 400- and 800-m street network buffers. Address data come from the 2008 Boston Youth Survey Geospatial Dataset, a school-based sample of public high school students in Boston, MA with complete residential address information (n = 1,292). GIS data were used to create multiple objective measures of neighborhood walkability and transit availability. We also obtained Walk Scores and Transit Scores. We calculated Spearman correlations of Walk Scores and Transit Scores with the GIS neighborhood walkability/transit availability measures as well as Spearman correlations accounting for spatial autocorrelation. Several significant correlations were observed between Walk Score and 400-m buffer GIS measures of neighborhood walkability; all significant correlations were found for the 800-m buffer. All correlations between Transit Scores and GIS measures of neighborhood transit availability were also significant (all p < 0. 0001). However, the magnitude of correlations varied by the GIS measure and neighborhood definition. Relative to the 400-m buffer, correlations for the 800-m buffer were higher. This study suggests that Walk Score is a good, convenient tool to measure certain aspects of neighborhood walkability and transit availability (such as density of retail destinations, density of recreational open space, intersection density, residential density and density of subway stops). However, Walk Score works best at larger spatial scales.

AB - We investigated the validity of Walk Scores and Transit Scores from the Walk Score website using several objective geographic information systems (GIS) measures of neighborhood walkabiltiy and transit availability based on 400- and 800-m street network buffers. Address data come from the 2008 Boston Youth Survey Geospatial Dataset, a school-based sample of public high school students in Boston, MA with complete residential address information (n = 1,292). GIS data were used to create multiple objective measures of neighborhood walkability and transit availability. We also obtained Walk Scores and Transit Scores. We calculated Spearman correlations of Walk Scores and Transit Scores with the GIS neighborhood walkability/transit availability measures as well as Spearman correlations accounting for spatial autocorrelation. Several significant correlations were observed between Walk Score and 400-m buffer GIS measures of neighborhood walkability; all significant correlations were found for the 800-m buffer. All correlations between Transit Scores and GIS measures of neighborhood transit availability were also significant (all p < 0. 0001). However, the magnitude of correlations varied by the GIS measure and neighborhood definition. Relative to the 400-m buffer, correlations for the 800-m buffer were higher. This study suggests that Walk Score is a good, convenient tool to measure certain aspects of neighborhood walkability and transit availability (such as density of retail destinations, density of recreational open space, intersection density, residential density and density of subway stops). However, Walk Score works best at larger spatial scales.

KW - Neighborhood walkability

KW - Small-area analysis

KW - Transit availability

KW - Transit score

KW - Validity

KW - Walk Score

UR - http://www.scopus.com/inward/record.url?scp=84876022535&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84876022535&partnerID=8YFLogxK

U2 - 10.1007/s10708-011-9444-4

DO - 10.1007/s10708-011-9444-4

M3 - Article

VL - 78

SP - 407

EP - 416

JO - Geo Journal

JF - Geo Journal

SN - 0343-2521

IS - 2

ER -