The Influence of Built Environment onWalking Behavior: Measurement Issues, Theoretical Considerations, Modeling Methodologies and Chinese Empirical Studies - Space-Time Integration in Geography and GIScience - page 65

Tabl e 5. 1 Typical empirical studies of walking behavior in China

Study
design

Walking
scenario Reference Sample Environmental attributes Scale Method Results

Pan et al.
( 2009 )
Shanghai

1,709 individuals
from four selected
neighborhoods
(conducted by
on-street
interviews)

CS, O (1) Distance to city center :

Distance to CBD and the
nearest city sub-center;
(2) Diversity of land use :
Commercial, industrial, etc.;
(3) Density : Population
density, average block size,
road length, road density;
(4) Distance to transit :
Number of bus stops and rail
transit stations within 1 km
radius of the neighborhood
centroid

Travel modal
choice:
Non-motorized
modes (walking
and cycling),
transit and
driving

N Multiple

logistic
regressions

In traditional
neighborhood with
relatively high densities
and small street blocks,
trip distance is shorter
and people are more
likely to choose walking
or cycling

Cervero and
Day ( 2008 )
Shanghai

A before-and-after
survey of 900
recent movers
(containing 2,840
inhabitants) in
three suburban
neighborhoods

LD, O
(relocation)

(1) Distance to transit :
Metro rail services, levels of
mixed-use activities;
(2) After relocation :
In superblock, less
walking-friendly, mixed-use
environments than residents'
previous neighborhoods

Mode change:
Non-motorized
to motorized
transport; bus to
other motorized
transport

R&N Descriptive

statistics and
binary logistic
models

Relocating to a
suburban area near a
metro-rail station
encourages commuters
to switch from
non-motorized and bus
transit to rail. Proximity
to regional rail networks
has stronger influence
on commuting behavior
than neighborhood
street designs and
land-use patterns
(continued)