Human Factors Assessment

of Pedestrian Roadway
Crossing Behavior
PUBLICATION NO. FHWA-HRT-13-098 JANUARY 2014

Research, Development, and Technology

Turner-Fairbank Highway Research Center
6300 Georgetown Pike
McLean, VA 22101-2296
 FOREWORD

More than half of pedestrian fatalities occur at unmarked locations away from intersections.
However, little research has been conducted to understand why pedestrians cross roadways at
unmarked locations. As a result, this study sought to better understand the environmental
influences on both where and when pedestrians elect to cross the road. This report examines
more than 70,000 pedestrian crossings at 20 different locations. The circumstances of those
crossings (pedestrians yielding to vehicles, vehicles yielding to pedestrians, and evasive actions)
were documented and analyzed. A model using environmental factors as inputs is provided to
predict where (marked crosswalk intersection or outside the marked crosswalk) pedestrians will
cross the road.

This report may be of interest to roadway designers, traffic engineers, and researchers who are
concerned with the safety of pedestrian crossings at unmarked locations.

Monique Evans
Director, Office of Safety
Research and Development

Notice
This document is disseminated under the sponsorship of the U.S. Department of Transportation
in the interest of information exchange. The U.S. Government assumes no liability for the use
of the information contained in this document. This report does not constitute a standard,
specification, or regulation.

The U.S. Government does not endorse products or manufacturers. Trademarks or

manufacturers’ names appear in this report only because they are considered essential to the
objective of the document.

Quality Assurance Statement

The Federal Highway Administration (FHWA) provides high-quality information to serve
Government, industry, and the public in a manner that promotes public understanding. Standards
and policies are used to ensure and maximize the quality, objectivity, utility, and integrity of its
information. FHWA periodically reviews quality issues and adjusts its programs and processes to
ensure continuous quality improvement.
 TECHNICAL DOCUMENTATION PAGE

1. Report No. 2. Government Accession No. 3. Recipient’s Catalog No.

FHWA-HRT-13-098
4. Title and Subtitle 5. Report Date
Human Factors Assessment of Pedestrian Roadway January 2014
Crossing Behavior 6. Performing Organization Code:

7. Author(s) 8. Performing Organization Report No.

Stacy A. Balk, Mary Anne Bertola, Jim Shurbutt, and Ann
Do
9. Performing Organization Name and Address 10. Work Unit No.
SAIC TRAIS
6300 Georgetown Pike
McLean, VA 22101-2296 11. Contract or Grant No.

12. Sponsoring Agency Name and Address 13. Type of Report and Period Covered
Federal Highway Administration
1200 New Jersey Ave. S.E.
14. Sponsoring Agency Code
Washington, DC 20590-9898

15. Supplementary Notes

The Contracting Officer’s Representative (COR) was C.Y. David Yang.
16. Abstract
Pedestrian–vehicle crashes are both common and deadly. The majority of pedestrian fatalities occur outside
marked intersection crosswalks. The influences of pedestrian and environmental factors on crossing location
choice were examined. A literature review covering factors intrinsic to pedestrians is provided. In addition,
pedestrian crossings at 20 different locations were recorded and analyzed. The vast majority of crossings
(89 percent of the total observed) took place in the marked intersection crosswalks. Drivers are likely to yield to
pedestrians. However, while drivers are more likely to yield to pedestrians in the marked crosswalk, pedestrians
and vehicles are equally as likely to yield to one another outside the marked crosswalk. The data also suggest that
measures that reduce the perceived affordances to cross the roadway (e.g., flowerbeds that separate the sidewalk
from the roadway) also reduce the proportion of crossings outside the marked crosswalks. It also appears that
pedestrians cross when perceived control of the crossing is greatest. Measures to increase perceived control have
the potential to increase (e.g., visible countdown clocks) or decrease (e.g., large medians) crossings in the marked
crosswalk. A model to predict pedestrian crossing location is provided. The model uses various environmental
variables as predicting factors and was shown to successfully predict an average of 90 percent of the crossings.
17. Key Words 18. Distribution Statement
Pedestrian, crosswalk, midblock crossings, safety, No restrictions. This document is available to the public
human factors assessment through NTIS: National Technical Information Service
5301 Shawnee Road, Alexandria, VA 22312
19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price
Unclassified Unclassified 218
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized
 SI* (MODERN METRIC) CONVERSION FACTORS
APPROXIMATE CONVERSIONS TO SI UNITS
Symbol When You Know Multiply By To Find Symbol
LENGTH
in inches 25.4 millimeters mm
ft feet 0.305 meters m
yd yards 0.914 meters m
mi miles 1.61 kilometers km
AREA
2 2
in square inches 645.2 square millimeters mm
2 2
ft square feet 0.093 square meters m
yd2 square yard 0.836 square meters m2
ac acres 0.405 hectares ha
2 2
mi square miles 2.59 square kilometers km
VOLUME
fl oz fluid ounces 29.57 milliliters mL
gal gallons 3.785 liters L
ft3 cubic feet 0.028 cubic meters m3
3 3
yd cubic yards 0.765 cubic meters m
3
NOTE: volumes greater than 1000 L shall be shown in m
MASS
oz ounces 28.35 grams g
lb pounds 0.454 kilograms kg
T short tons (2000 lb) 0.907 megagrams (or "metric ton") Mg (or "t")
TEMPERATURE (exact degrees)
o o
F Fahrenheit 5 (F-32)/9 Celsius C
or (F-32)/1.8
ILLUMINATION
fc foot-candles 10.76 lux lx
2 2
fl foot-Lamberts 3.426 candela/m cd/m
FORCE and PRESSURE or STRESS
lbf poundforce 4.45 newtons N
lbf/in2 poundforce per square inch 6.89 kilopascals kPa

APPROXIMATE CONVERSIONS FROM SI UNITS

Symbol When You Know Multiply By To Find Symbol
LENGTH
mm millimeters 0.039 inches in
m meters 3.28 feet ft
m meters 1.09 yards yd
km kilometers 0.621 miles mi
AREA
mm2 square millimeters 0.0016 square inches in2
2 2
m square meters 10.764 square feet ft
2 2
m square meters 1.195 square yards yd
ha hectares 2.47 acres ac
2 2
km square kilometers 0.386 square miles mi
VOLUME
mL milliliters 0.034 fluid ounces fl oz
L liters 0.264 gallons gal
m3 cubic meters 35.314 cubic feet ft3
3 3
m cubic meters 1.307 cubic yards yd
MASS
g grams 0.035 ounces oz
kg kilograms 2.202 pounds lb
Mg (or "t") megagrams (or "metric ton") 1.103 short tons (2000 lb) T
TEMPERATURE (exact degrees)
o o
C Celsius 1.8C+32 Fahrenheit F
ILLUMINATION
lx lux 0.0929 foot-candles fc
2 2
cd/m candela/m 0.2919 foot-Lamberts fl
FORCE and PRESSURE or STRESS
N newtons 0.225 poundforce lbf
2
kPa kilopascals 0.145 poundforce per square inch lbf/in
*SI is the symbol for the International System of Units. Appropriate rounding should be made to comply with Section 4 of ASTM E380.
(Revised March 2003)

ii
 TABLE OF CONTENTS

EXECUTIVE SUMMARY ...........................................................................................................1

INTRODUCTION..........................................................................................................................5
PEDESTRIAN FACTORS ......................................................................................................7
Gender ..................................................................................................................................7
Age .......................................................................................................................................7
Alcohol.................................................................................................................................8
Self-Identity .........................................................................................................................8
Perceived Control.................................................................................................................9
ENVIRONMENTAL FACTORS ...........................................................................................9
Trip Originators .................................................................................................................10
Trip Destinations................................................................................................................10
Affordances ........................................................................................................................10
PEDESTRIAN CROSSING BEHAVIORS: PHASE 1 ............................................................13
DATA COLLECTION LOCATIONS .................................................................................13
Location 1 ..........................................................................................................................13
Location 2 ..........................................................................................................................16
Location 3 ..........................................................................................................................19
Location 4 ..........................................................................................................................22
Location 5 ..........................................................................................................................24
Location 6 ..........................................................................................................................27
Location 7 ..........................................................................................................................30
Location 8 ..........................................................................................................................32
DATA COLLECTION VALIDATION ...............................................................................35
VIDEO DATA CODING .......................................................................................................36
Crossing Factors.................................................................................................................36
Dates/Times Coded ............................................................................................................38
Inter-Rater Reliability ........................................................................................................39
RESULTS ...............................................................................................................................39
Descriptive Data.................................................................................................................39
Predictive Model ................................................................................................................65
Factor Specific Analyses....................................................................................................70
DISCUSSION .........................................................................................................................79
Crossing Location ..............................................................................................................79
Pedestrian Yielding ............................................................................................................82
Vehicle Yielding ................................................................................................................83
Yielding Comparison .........................................................................................................83
Evasive Pedestrian Actions ................................................................................................83
Evasive Vehicle Actions ....................................................................................................84
Evasive Action Comparison ..............................................................................................84
General Discussion ............................................................................................................84
PEDESTRIAN CROSSING BEHAVIORS: PHASE 2 ............................................................85
DATA COLLECTION LOCATIONS .................................................................................85
Location 9 ..........................................................................................................................85

iii
 Location 10 ........................................................................................................................87
Location 11 ........................................................................................................................90
Location 12 ........................................................................................................................92
Location 13 ........................................................................................................................95
Location 14 ........................................................................................................................97
Location 15 ......................................................................................................................100
DATA COLLECTION VALIDATION .............................................................................102
VIDEO DATA CODING .....................................................................................................103
Crossing Factors...............................................................................................................103
Dates/Times Coded ..........................................................................................................103
RESULTS .............................................................................................................................104
Descriptive Data...............................................................................................................104
Predictive Model ..............................................................................................................120
Factor-Specific Analyses .................................................................................................120
DISCUSSION .......................................................................................................................125
Crossing Location ............................................................................................................126
Pedestrian Yielding ..........................................................................................................127
Vehicle Yielding ..............................................................................................................128
Yielding Comparison .......................................................................................................128
Evasive Pedestrian Actions ..............................................................................................128
Evasive Vehicle Actions ..................................................................................................128
General Discussion ..........................................................................................................128
PEDESTRIAN CROSSING BEHAVIORS: PHASE 3 ..........................................................131
DATA COLLECTION LOCATIONS ...............................................................................131
Location 16 ......................................................................................................................131
Location 17 ......................................................................................................................134
Location 18 ......................................................................................................................136
Location 19 ......................................................................................................................138
Location 20 ......................................................................................................................142
DATA CODING ...................................................................................................................144
Crossing Factors...............................................................................................................144
Dates/Times Coded ..........................................................................................................144
RESULTS .............................................................................................................................145
Descriptive Data...............................................................................................................145
Predictive Model ..............................................................................................................157
Factor-Specific Analyses .................................................................................................157
DISCUSSION .......................................................................................................................162
PEDESTRIAN CROSSING BEHAVIORS: COMBINED RESULTS .................................163
Full Dataset ......................................................................................................................166
Crossing Proportions........................................................................................................176
DISCUSSION .......................................................................................................................190
Pedestrian Factors ............................................................................................................190
Environmental Factors .....................................................................................................192
Summary ..........................................................................................................................201
REFERENCES ...........................................................................................................................203

iv
 LIST OF FIGURES

Figure 1. Graph. Total number of pedestrian fatalities, number of urban intersection/

intersection-related fatalities, and number of urban non-intersection fatalities in
daytime environments by year .............................................................................................6
Figure 2. Photo. Aerial view of Location 1 ...................................................................................14
Figure 3. Photo. Still image captured from the Location 1 camera ...............................................15
Figure 4. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side of H Street Northeast to the south side of the street along 3rd
Street Northeast..................................................................................................................16
Figure 5. Photo. Aerial view of Location 2 ...................................................................................17
Figure 6. Photo. Still image captured from the Location 2 camera ...............................................18
Figure 7. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side of New York Avenue Northwest to the south side of the street
along 7th Street Northwest.................................................................................................19
Figure 8. Photo. Aerial view of Location 3. Northwest between 7th and 6th Streets
Northwest ...........................................................................................................................20
Figure 9. Photo. Still image captured from the Location 3 camera ...............................................21
Figure 10. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side of Rhode Island Avenue Northwest to the south side of the street
along 7th Street Northwest.................................................................................................22
Figure 11. Photo. Aerial view of Location 4 .................................................................................23
Figure 12. Photo. Still image captured from the Location 4 camera .............................................23
Figure 13. Photo. Image of the westernmost crossing of I Street Northwest at 13th Street
Northwest. The photo is taken from the southwestern corner of the intersection .............24
Figure 14. Photo. Aerial view of Location 5 .................................................................................25
Figure 15. Photo. Still image captured from the Location 5 camera .............................................26
Figure 16. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the south side of H Street Northwest to the north side of the street along
14th Street Northwest ........................................................................................................27
Figure 17. Photo. Aerial view of Location 6 .................................................................................28
Figure 18. Photo. Still image captured from the Location 6 camera .............................................29
Figure 19. Photo. An image of the southernmost crossing of 14th Street Northwest at New
York Avenue Northwest. The photo is taken from the northwestern corner of the
intersection .........................................................................................................................30
Figure 20. Photo. Aerial view of Location 7 .................................................................................31
Figure 21. Photo. Still image captured from the Location 7 camera .............................................31
Figure 22. Photo. Illustration of what a pedestrian might see as he or she attempts to crossing
from the west side of 19th Street Northwest to the east side of the street along
Pennsylvania Avenue Northwest .......................................................................................32
Figure 23. Photo. Aerial view of Location 8 .................................................................................33
Figure 24. Photo. Still image captured from the Location 8 camera .............................................34

v
Figure 25. Photo. An image of the (far) flag crossing at Connecticut Avenue Northwest and
Northampton Street Northwest ..........................................................................................34
Figure 26. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the west side of Connecticut Avenue Northwest to the east side of the street
along Oliver Street Northwest ...........................................................................................35
Figure 27. Photo. Aerial view of Location 9 .................................................................................86
Figure 28. Photo. Still image captured from the Location 9 camera .............................................86
Figure 29. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the west side of Connecticut Avenue Northwest to the east side of the street
along Florida Avenue Northwest .......................................................................................87
Figure 30. Photo. Aerial view of Location 10 ...............................................................................88
Figure 31. Photo. Still image captured from the Location 10 camera ...........................................89
Figure 32. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the west side of Connecticut Avenue Northwest to the east side of the street
along Van Ness Street Northwest ......................................................................................90
Figure 33. Photo. Aerial view of Location 11 ...............................................................................91
Figure 34. Photo. Still image captured from the Location 11 camera ...........................................91
Figure 35. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the west side of Georgia Avenue Northwest to the east side of the street along
Arkansas Avenue Northwest..............................................................................................92
Figure 36. Photo. Aerial view of Location 12 ...............................................................................93
Figure 37. Photo. Still image captured from the Location 12 camera ...........................................94
Figure 38. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the east side of Georgia Avenue Northwest to the west side of the street along
Irving Street Northwest ......................................................................................................95
Figure 39. Photo. Aerial view of Location 13 ...............................................................................96
Figure 40. Photo. Still image captured from the Location 13 camera ...........................................96
Figure 41. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side of H Street Northwest to the south side of the street along
Connecticut Avenue Northwest .........................................................................................97
Figure 42. Photo. Aerial view of Location 14 ...............................................................................98
Figure 43. Photo. Still image captured from the Location 14 camera ...........................................99
Figure 44. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side of Independence Avenue Southwest to the south side of the
street along Washington Avenue Southwest....................................................................100
Figure 45. Photo. Aerial view of Location 15 .............................................................................101
Figure 46. Photo. Still image captured from the Location 15 camera .........................................101
Figure 47. Photo. Illustration of what a pedestrian might see as he or she attempts to make a
crossing from the south side of Pennsylvania Avenue Southeast to the north side of
the street along 8th Street Southeast ................................................................................102
Figure 48. Photo. Aerial view of Location 16 .............................................................................132
Figure 49. Photo. Illustration of some of the foliage present in the median of North
Washington Street ............................................................................................................133

vi
Figure 50. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the east side to the west side of North Washington Street along King Street .........134
Figure 51. Photo. Aerial view of Location 17 .............................................................................135
Figure 52. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the south side to the north side of King Street along North Washington Street .....136
Figure 53. Photo. Aerial view of location 18 ...............................................................................137
Figure 54. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side to the south side of Irving Street Northwest along 14th Street
Northwest in Washington DC ..........................................................................................138
Figure 55. Photo. Aerial view of location 19 ...............................................................................139
Figure 56. Photo. View of the planters located between the southbound lanes of traffic and
the adjacent sidewalk along Wisconsin Avenue ..............................................................140
Figure 57. Photo. Illustration of some of the foliage and elevation present in the median of
Wisconsin Avenue ...........................................................................................................141
Figure 58. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the west side to the east side of Wisconsin Avenue along Willard Avenue ...........142
Figure 59. Photo. Aerial view of Location 20 .............................................................................143
Figure 60. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the south side to the north side of Clarendon Boulevard along North Edgewood
Street ................................................................................................................................144
Figure 61. Chart. Distribution of all crossings observed across the 20 different locations by
the area in which they took place.....................................................................................166
Figure 62. Chart. Distribution of crossings observed across all 20 locations in the marked
intersections by circumstances of the crossing ................................................................167
Figure 63. Chart. Distribution of crossings observed across all 20 locations in the unmarked
non-intersections by the circumstances of the crossing ...................................................167
Figure 64. Chart. Distribution of crossings observed across all 20 locations for all rule-
breaking crossings by the circumstances of the crossing (including those in the
unmarked non-intersections)............................................................................................168
Figure 65. Chart. Distribution of crossings observed across all 20 locations for crossings
made entirely during the walk light phase in the marked intersections, by the
circumstances of the crossing ..........................................................................................168
Figure 66. Chart. Distribution of crossings observed across all 20 locations for rule-breaking
crossings in marked intersections (i.e., crossings that took place at least partially
during the don’t walk light phase) by the circumstances of the crossing ........................169
Figure 67. Chart. Scatterplot of each location’s score for each of the five factors against the
percentage of intersection crossings at that location........................................................173
Figure 68. Equation. The logit of crossing at the intersection .....................................................173
Figure 69. Chart. Scatterplot of each location’s score for each of the five factors against the
percentage of intersection crossings at that location........................................................175
Figure 70. Equation. The logit of crossing at the intersection during the walk phase .................175

vii
 LIST OF TABLES

Table 1. Relevant sunrise, sunset, and weather for each of the coded data collection days ..........39
Table 2. Summary of pedestrian crossings combined across all eight data collection
locations .............................................................................................................................40
Table 3. Summary of pedestrian crossings at Location 1 ..............................................................43
Table 4. Summary of pedestrian crossings at Location 2 ..............................................................46
Table 5. Summary of pedestrian crossings at Location 3 ..............................................................49
Table 6. Summary of pedestrian crossings at Location 4 ..............................................................52
Table 7. Summary of pedestrian crossings at Location 5 ..............................................................55
Table 8. Summary of pedestrian crossings at Location 6 ..............................................................58
Table 9. Summary of pedestrian crossings at Location 7 ..............................................................61
Table 10. Summary of pedestrian crossings at Location 8 ............................................................64
Table 11. Predictors and their respective descriptions used for the models ..................................66
Table 12. Summary of each of the model-selected effects (predictors) and their parameter
estimates.............................................................................................................................70
Table 13. Percentage of pedestrians at each crossing area in each data collection location..........72
Table 14. Percentage of pedestrians yielding to vehicles in each crossing area at each data
collection location ..............................................................................................................73
Table 15. Percentage of vehicles yielding to pedestrians within each crossing area at each
data collection location ......................................................................................................75
Table 16. Percentage of pedestrian evasive actions in each crossing area at each data
collection location ..............................................................................................................77
Table 17. Percentage of vehicle evasive actions within each crossing area at each data
collection location ..............................................................................................................78
Table 18. Relevant sunrise, sunset, and weather for each of the coded data collection days ......104
Table 19. Summary of pedestrian crossings—combined across all seven data collection
locations ...........................................................................................................................105
Table 20. Summary of pedestrian crossings at Location 9 ..........................................................107
Table 21. Summary of pedestrian crossings at Location 10 ........................................................109
Table 22. Summary of pedestrian crossings at Location 11 ........................................................111
Table 23. Summary of pedestrian crossings at Location 12 ........................................................113
Table 24. Summary of pedestrian crossings at Location 13 ........................................................115
Table 25. Summary of pedestrian crossings at Location 14 ........................................................117
Table 26. Summary of pedestrian crossings at Location 15 ........................................................119
Table 27. Percentage of pedestrians at each crossing area in each data collection location........120
Table 28. Percentage of pedestrians yielding to vehicles in each crossing area at each data
collection location ............................................................................................................121
Table 29. Percentage of vehicles yielding to pedestrians within each crossing area at each
data collection location ....................................................................................................123
Table 30. Percentage of pedestrian evasive actions within each crossing area at each data
collection location ............................................................................................................124

viii
Table 31. Percentage of vehicle evasive actions within each crossing area at each data
collection location ............................................................................................................125
Table 32. Relevant sunrise, sunset, and weather for each of the coded data collection days ......145
Table 33. Summary of pedestrian crossings—combined across data collection locations
16–20................................................................................................................................146
Table 34. Summary of pedestrian crossings at Location 16 ........................................................148
Table 35. Summary of pedestrian crossings at Location 17 ........................................................150
Table 36. Summary of pedestrian crossings at Location 18 ........................................................152
Table 37. Summary of pedestrian crossings at Location 19 ........................................................154
Table 38. Summary of pedestrian crossings at Location 20 ........................................................156
Table 39. Percentage of pedestrians at each crossing area in each data collection location........157
Table 40. Percentage of pedestrians yielding to vehicles in each crossing area at each data
collection location ............................................................................................................158
Table 41. Percentage of vehicles yielding to pedestrians in each crossing area at each data
collection location ............................................................................................................159
Table 42. Percentage of pedestrian evasive actions in each crossing area at each data
collection location ............................................................................................................161
Table 43. Percentage of vehicle evasive actions in each crossing area at each data collection
location.............................................................................................................................162
Table 44. Summary of pedestrian crossing locations and their associated environmental
factors that were assessed on a continuous scale .............................................................164
Table 45. Summary of pedestrian crossing locations and their associated environmental
factors that were assessed on a categorical scale .............................................................165
Table 46. Environmental factors and their labels used to calculate the crossing location
prediction model ..............................................................................................................170
Table 47. Factor loadings for the 16 environmental variables.....................................................171
Table 48. Standardized scoring coefficients for the five rotated factors .....................................172
Table 49. Probabilities of correctly predicting a crossing at the marked intersection or the
unmarked non-intersection by location and the corresponding upper and lower 95-
percent confidence limits .................................................................................................174
Table 50. Probabilities of correctly predicting a crossing at the marked intersection entirely
during the walk phase by location and the corresponding upper and lower 95-percent
confidence limits ..............................................................................................................176
Table 51. Percentage of pedestrians crossing at each area in each data collection location........178
Table 52. Percentage of pedestrians yielding to vehicles in each crossing area at each data
collection location ............................................................................................................181
Table 53. Percentage of vehicles yielding to pedestrians in each crossing area at each data
collection location ............................................................................................................184
Table 54. Percentage of pedestrian evasive actions in each crossing area at each data
collection location ............................................................................................................187
Table 55. Percentage of vehicle evasive actions in each crossing area at each data collection
location.............................................................................................................................189

ix
 EXECUTIVE SUMMARY

Pedestrian–vehicle crashes are both common and deadly. In 2010, 13 percent of all fatal crashes
involved pedestrians.(1) Of these, 68.1 percent occurred outside intersections. As a result of the
large proportion of pedestrian fatalities that occur at non-intersection locations, it is important to
investigate the causal factors of these collisions. Despite the large proportion of crashes, little
research has investigated the reasons pedestrians cross roadways at unmarked locations.

As a result, the present study sought to better understand the environmental influences on both
where and when pedestrians elect to cross the road. The circumstances surrounding when and
where more than 70,000 crossings took place were recorded and analyzed. A model to predict
crossing behaviors was created. These data have the potential to guide roadway design.
Furthermore, this approach may aid in the selection and location of pedestrian crossing
interventions (e.g., new pedestrian activation crossing beacons), ultimately increasing pedestrian
safety in shared use environments.

Pedestrian roadway crossings were coded at 20 different locations in the Washington, DC,
metropolitan area. Each location was one block in length and was flanked by two marked
crosswalks at intersections. Crossings were recorded within one marked, light-controlled
crosswalk and the roadway between it and the next marked crossing (but not within the far
crossing). Pedestrian crossings were coded for several different factors:

A. Location. Within the marked crosswalk, or not.

B. Traffic status. Walk or don’t walk sign illuminated.

C. Yielding. Pedestrians yielding to vehicles or vehicles yielding to pedestrians in the

roadway.

D. Evasive Actions. Any evasive movement made by a vehicle or pedestrian to avoid

collision (e.g., running or abrupt braking).

Stable components of each location were also recorded:

1. Distance between the marked crosswalks.

2. Average annual daily traffic volume (AADT).

3. Street directionality (one- or two-way).

4. Physical barriers in or along the roadway that might prevent pedestrians from easily
crossing between the roadway and sidewalk.

5. Presence and location of bus stops.

6. Number of potential pedestrian trip originators/destinations.

7. Availability of street parking.

1
 8. Presence of a center turn lane.

9. Presence of a right turn only turning lane

10. Length of the walk light phase.

11. Length of the don’t walk light phase.

12. Width of the roadway/pedestrian crossing.

13. Presence and type of median (e.g., raised concrete or painted asphalt).

14. Presence of a T-intersection between the two marked crosswalks.

15. Traffic control device of the second crosswalk (i.e., traffic signal, stop sign, or none).

16. Pace at which pedestrians are required to travel to complete a crossing entirely during the
walk light phase.

Data were used to create a model to predict where pedestrians are likely to cross the road
(marked intersection crosswalk or non-intersection). The accuracy of the model ranged from
80 to 95 percent based on location. The model correctly predicted a mean of 90 percent of
crossings. Overall, the model was successful in predicting whether pedestrians would cross in
marked crosswalks at intersections or outside a marked crossing.

A mean of 13.89 percent of pedestrian crossings took place at unmarked non-intersection

locations. Given the disproportionate percentage of fatalities that take place outside marked
intersections, this number may be a bit surprising. However, these data suggest that some
locations are more prone to have more unmarked non-intersection crossings than others. This
was indeed the case here. Non-intersection crossings ranged from 3.02 to 36.55 percent.

The location with 36.55 percent of the crossings that took place at a non-intersection was
different from many of the other locations in very specific ways. A wide, grassy median
separates traffic directionality. This median allows pedestrians to cross one road segment, wait
on the median for a gap in traffic, and complete the second portion of the crossing. In addition,
the juxtaposition of a Metro (subway) train station and a surrounding neighborhood is such that
the most direct route (in terms of absolute distance) between the two areas involves crossing
outside the marked intersection. Given that some might consider traveling through the marked
crosswalks to be out of the way, many may increase their perceived control of the crossing by
using the median and cross midblock.

Environmental factors were also examined in terms of their influence on crossing behaviors. For
example, a significant relationship between the width of the crossing and the percentage of
pedestrians who crossed entirely during the walk signal light phase at each location was found.
In other words, the longer the distance that pedestrians were required to travel to cross the road,
the more likely they were to cross entirely during the walk phase of the light cycle. Interestingly,
a significant relationship between crossing entirely during the don’t walk signal phase and traffic

2
directionality was found—pedestrians were more likely to cross during the don’t walk phase on
one-way streets than on two-way streets.

Not surprisingly, when physical barriers that might prevent pedestrians from easily crossing
between the roadway and sidewalk were present, pedestrians were less likely to cross the
roadway at unmarked non-intersection areas. Thus it appears that even small barriers, such as
flower planters, reduce the perceived affordances to cross the roadway.

Overall, only .98 percent of crossings involved pedestrians yielding to vehicles. Not surprisingly,
a significantly greater percentage of crossings in non-intersections involved pedestrian yielding
than in marked crosswalks.

Overall, 8.93 percent of crossings involved a vehicle yielding to a pedestrian. A significantly

greater percentage of crossings in the marked intersection involved vehicle yielding than
crossings in the unmarked non-intersection areas. This discrepancy is largely attributable to
turning vehicles yielding to pedestrians crossing in the marked crossings during the walk phase.

Within the marked intersections, a significantly greater percentage of crossings involved vehicle
yielding than pedestrian yielding. However, outside the marked pedestrian crossing, pedestrians
and vehicles were equally likely to yield to avoid collision.

It is recommended that in new, redesigned, or problematic environments an evaluation of the

environmental features should be made to determine where pedestrian crossings are likely. The
developed model was successful in predicting an average of 90 percent of the pedestrian
crossings. Areas that have a high predicted likelihood of unmarked non-intersection crossings
could be proactively targeted to modify the crossing affordances of the environment—leading
pedestrians to cross at marked intersections. Presumably this would reduce the number of
pedestrians crossing midblock. A combined effort of pedestrian education and shared road use
planning would hopefully reduce pedestrian injuries and fatalities and ultimately increase
roadway safety.

3
 INTRODUCTION

Vehicle collisions with pedestrians are both common and deadly. In 2009, 12.1 percent (a
number that has remained relatively consistent since 2000, ranging from 12.1 percent to
12.9 percent) of roadway fatalities in the United States were pedestrians.(1) The problem is even
graver elsewhere in the world. For example, 67 percent of roadway fatalities in Hong Kong,
51 percent in Ethiopia, and 50 percent in Romania are pedestrians.(2) It has been estimated that
approximately 400,000 pedestrian fatalities occur globally each year.(3) Further, the World
Health Organization estimates that roadway crashes will become the third leading cause of years
of life lost by 2020. (Years of life lost is a measure quantifying premature mortality.)(4)

An examination of pedestrian fatalities in the United States reveals that most occur at non-
intersection locations.(1) As a result, it is important both to be able to determine where these
collisions are likely to occur and to understand what measures can be taken to improve safety.
The present study focuses on the former. In other words, this study focuses on determining what
factors, both pedestrian and environmental, influence potentially dangerous non-intersection
crossing behavior. It is hoped that this information might be used to increase the proportion of
safe pedestrian crossing behaviors.

During the daytime in the United States, about 1,500 pedestrian fatalities occur in urban
environments each year (based on an analysis of Fatality Analysis Reporting System (FARS)
data for 1994–2009).(1) Of these, about 64 percent occur on the roadway in non-intersection
locations, while only about 30 percent occur in intersections or are intersection related.

Figure 1 illustrates this relationship. It should be noted that the number of non-intersection and
intersection/intersection-related fatalities do not equate to the total number of daytime fatalities.
This is because areas such as driveways, exit ramps, and rail grade crossings were excluded from
both the intersection/intersection-related and non-intersection categories. (These areas were
excluded from analyses because they result in special circumstances that are distinctly different
from traditional non-intersection crossing.) In addition, rural environments, interstates,
expressways, and unknown roadway types were excluded from the non-intersection and
intersection/intersection-related categories. These areas were excluded from analyses because of
the focus of the current study. This study focuses on urban areas where pedestrian populations
are the densest. Furthermore, these urban areas have the greatest likelihood of crossing treatment
safety enhancements (e.g., pedestrian hybrid beacon). In addition, as a result of the perceptual
complexities involved with nighttime pedestrian collisions, the present study focuses only on
those occurring during daylight hours.

As a result of the large proportion of pedestrian fatalities that occur at non-intersection locations,
it is important to investigate the causal factors of these collisions. It is probable that some of
these collisions result from pedestrians traveling along the roadway. However, it is likely that the
majority result from pedestrians crossing the roadway. Despite the large proportion of crashes
involving pedestrians crossing roadways at unmarked locations, little research has investigated
the reasons.

5
 Figure 1. Graph. Total number of pedestrian fatalities, number of urban intersection/
intersection-related fatalities, and number of urban non-intersection fatalities in daytime
environments by year.

It has been shown that pedestrians who cross away from crosswalks are more likely to be trapped
in the middle of the street than those who cross at the crosswalk.(5) Furthermore, pedestrians who
cross at crosswalks are more likely to yield to vehicles than those who cross away from
crosswalks.(5) This is especially important to safety. If pedestrians take measures to increase
their safety, they are less likely to be involved in pedestrian–vehicle collisions.

In many areas, pedestrian markings and signalization can be implemented to attempt to increase
safety as pedestrians cross midblock. However, it is obvious that some areas simply do not
feasibly (or realistically) allow for the addition of marked pedestrian crossings (e.g., rural and
gridlock-prone roads).

The Federal Highway Administration (FHWA) has multiple current projects investigating the
performance of several different pedestrian crossing treatments. These projects include
identifying the optimal number and combination of rectangular rapid flashing beacons and
circular rapid flashing beacons, as well as examining the effect of varying characteristics, such as
flash pattern/sequence, intensity, size, shape, and positioning.

No matter the situation, it is important to understand what types of factors influence pedestrians
to cross at unmarked locations. Roadway designs and environmental factors have the ability to
both promote and hinder opportunities to cross at unmarked and midblock locations. For
example, a concrete lane divider severely reduces the ability to quickly and safely cross a
roadway. Yet, some areas appear to encourage people to cross at unmarked locations. Pedestrians
may elect to cross the road for many reasons. This report focuses on two of these factors—those
intrinsic to pedestrians and those of the roadway environment itself.

6
PEDESTRIAN FACTORS

It is obvious that pedestrians, in most cases, attempt to select the fastest and most direct route
from their origin to their preferred destination. However, route planning must take into account
the features of the traveling environment, physical abilities, some level of risk (real or
perceived), among other factors. For example, suppose that Route A is .5 miles in length and that
Route B is 1 mi in length. While Route A is shorter in absolute distance, it includes two steep hill
segments, and Route B is flat. Given these environmental characteristics, a young and physically
fit pedestrian would likely select Route A. However, an older adult with a physical impairment
might select Route B for its terrain advantages.

It is improbable that people use complex decisionmaking processes to determine which route to
walk to reach a specific location. Rather, pedestrians are likely to plan a general route and rely
on other factors to make smaller in-route choices, such as where to cross the roadway. For
example, one person may “always” cross at a specific location, and for this reason, he or she will
rely on habit and cross there again on subsequent trips. Another pedestrian may have mobility
issues and will cross at the most accessible location. It is probable that pedestrians rely on a set
of heuristics or tendencies when selecting when and where to cross the roadway. Here, the study
team explores some of the intrinsic pedestrian characteristics that can influence route choice and
crossing behaviors.

Gender

Overall, males are more likely to be involved in vehicle crashes than females.(6) This trend
remains consistent with pedestrian fatalities. In 2009, the male pedestrian fatality rate was
1.86 per 100,000, compared with .82 per 100,000 females. In total, 69 percent of the pedestrian
fatalities were male. This gender difference is consistent across all age groups.(7)

This evidence suggests that gender plays a role in pedestrian crossing decisions to some degree.
This idea is further supported in a recent study by Holland and Hill.(8) The authors presented
participants with a variety of potentially dangerous crossing scenarios. In each of the scenarios,
women reported perceiving more risk than their male counterparts. Furthermore, the females
were less likely to indicate an intention to cross the roadway. Taken together, these data suggest
that, in general, males and females view and act on potential safety hazards in the roadway
environment differently. These gender differences might prove useful in targeting interventions
and education toward different groups.

Age

Age also plays a role in pedestrian–vehicle collisions. In 2009, pedestrians 65 and older had the
highest rate of fatalities, 1.96 per 100,000. Moreover, older adults constituted 19 percent of all
pedestrian fatalities.(7) On the opposite end of the spectrum, pedestrians 15 years and younger
accounted for 25 percent of all pedestrian injuries and 7 percent of pedestrian fatalities in
2009.(7)

Many factors could result in these age-related fatality trends. One possibility is that younger and
older pedestrians make riskier road-crossing choices. However, multiple studies have shown that
when participants are presented with potentially dangerous crossing scenarios, older adults are

7
less likely to indicate that they would be likely to attempt the crossing.(8,9,10) Thus, it appears, at
least in regard to intention to cross the road, older adults are not necessarily making more risky
choices. In fact, this suggests just the opposite. One can logically draw the conclusion that those
who are less likely to engage in risky crossing behaviors are also less likely to be involved in
pedestrian–vehicle collisions.

If older adults are less inclined than younger adults to intend to make risky crossings, then other
factors must contribute to their large proportion of the total pedestrian fatalities. This appears to
be the case. While older adults constitute 19 percent of all pedestrian fatalities, they only make
up 8.5 percent of the total injuries.(7) These data imply that older adults are less likely to be
struck by a vehicle. However, when an older adult is involved in this type of collision, it is more
likely to result in a fatality.

It has been suggested, for example in Barton and Morrongiello, that children have not fully
developed the cognitive reasoning skills to cross the street safely.(11) It is certainly a possibility
that children make riskier crossing decisions. It is also likely that smaller children are more
difficult to see from the drivers’ perspective. This condition makes it more difficult for a driver
to proactively avoid collision with this special group of pedestrians.

Age differences in pedestrian injury and fatality rates may help to direct educational
interventions to increase safe crossing behaviors. Further, age-related differences may help
engineers to target different interventions for different age groups. For example, sidewalk
markings outside elementary schools may lead children to walk to pedestrian-activated
crosswalks. In addition, crosswalk activation buttons could be lowered to child level or made
visually more attractive to push. Longer protected crossings could also be employed near senior
citizen communities.

Alcohol

Alcohol plays a significant role in pedestrian fatalities. In 2009, it was reported that 48 percent of
crashes resulting in a pedestrian fatality involved alcohol. When focusing on drivers involved in
these collisions, about 13 percent had a blood alcohol concentration (BAC) of .08 percent by
volume or greater. This is compared with 35 percent of pedestrians. Further, both the driver and
the pedestrian had BAC levels of .08 percent by volume or higher in 6 percent of the fatal
crashes.(7)

Self-Identity

Self-identity as a safe or careful pedestrian has been shown to be correlated with roadway
crossing behavior. For example, Holland, Hill, and Cooke asked pedestrians to make decisions
on whether or not they would cross the roadway in various situations.(10) It was found that
pedestrians who had high self-identities as careful pedestrians selected fewer gaps in which they
would cross the roadway at non-signalized crossing locations. Similarly, Evans and Norman
presented pedestrians with a set of potentially hazardous road crossing behaviors. Those people
who identified themselves as “safe pedestrians” were less likely to state that they would cross the
roadway in such a manner.(9) A similar study with adolescents also found that those who rated

8
themselves as safe pedestrians reported that they were less inclined to cross at a potentially risky
location.(12)

If people who identify themselves as cautious accept fewer gaps to cross the roadway, then it is
ideal to further understand what factors influence this self-identity. Holland, Hill, and Cooke
found that people who had previously been involved in a roadway collision viewed themselves
as more willing to take a risk.(10) Contrarily, those with mobility impairments and older adults
rated themselves as more cautious. Furthermore, it was found that the biggest predictor of a self-
identity as a careful or safe pedestrian was age (even when accounting for factors such as
experience and mobility). This suggests that older adults take fewer risks when crossing the
roadway than do younger pedestrians. Furthermore, this helps to explain the age-related
discrepancy in the number of collisions with younger persons than older persons. Beyond this, if
it is known that older adults are pedestrians who take fewer risks, then efforts to improve
roadway safety might be better directed at younger foot travelers.

The results of these studies suggest that people who consider themselves to be careful/safe
pedestrians are more aware of the potential hazards involved in crossing the roadway. As a
result, these people may be less likely to employ potentially risky road-crossing choices. Along
similar lines, it may be reasonable to assume that increases in perceived roadway dangers (either
via environmental changes or educational interventions) will lead to fewer crossings at unmarked
roadway locations.

Perceived Control

In addition to the aforementioned factors, Evans and Norman found that perceived control
influenced pedestrians’ intentions to cross the roadway.(12) (Perceived behavioral control is a
component of the Theory of Planned Behavior, a widely used theory of safety- and health-related
behaviors.(13)) Participants were presented with three scenarios: crossing during a gap in traffic of
a multilane roadway to get to a vehicle after a shopping trip, crossing at signalized pedestrian
crossing during the don’t walk phase to get to the dry cleaners before closing, and crossing a
residential roadway to meet a friend. For each of the scenarios, it was probable that those
participants who perceived having more behavioral control (as assessed by a planned behavior
questionnaire) of the situation would indicate that they would cross the roadway.

Evans and Norman also attempted to apply the Theory of Planned Behavior to adolescent
intentions to cross the roadway.(12) Middle school aged children were presented with a potentially
dangerous road-crossing scenario. Much like the researchers’ previously mentioned findings,
perceived behavioral control was the strongest predictor of expressed intention to cross the
roadway.(9) In both studies, perceived control of the situation is the greatest predictor of
pedestrians’ planned intent to cross the roadway. This suggests that if pedestrians’ perceptions of
the ability to cross the roadway at unsafe areas are decreased, then overall pedestrian safety may
increase.

ENVIRONMENTAL FACTORS

Many factors contribute to whether people attempt to cross the roadway at a specific location at a
specific time. However, features of the roadway environment have been largely ignored when

9
examining these causal factors. Here the study team discusses foot travel in three components:
trip originators, destinations or attractions, and affordances.

Trip Originators

Trip originators are areas where pedestrians begin trips. Some originators tend to generate more
trips than others. For example, a house generates a finite number of trips. However, other
originators generate countless pedestrian trips. Some of these high trip generating sources are
places such as shopping malls, Metro/subway stations, and bus stops. These high trip originators
require special attention with regard to pedestrian flow. One can easily imagine how pedestrian
travel patterns might differ between a bus stop placed in the middle of a block and a bus stop
placed at the corner of an intersection.

Trip Destinations

Trip destinations are end points of pedestrian trips, whether they are final destinations or
attractions (e.g., coffee shop) en route. Often, destinations are also trip originators. Take the
example of a shopping center. Many people may consider the shopping center the completion of
their trip, that is, their destination. However, when people leave this locale, the shopping center
becomes the trip originator. As a result, designers must attend to both how pedestrians enter and
exit such locations.

Affordances

It is obvious from the pedestrian fatality data that simple risk perception is not an adequate
source of determining whether or not one should cross the roadway at a given location at a given
time. If risk perceptions were adequate, pedestrians would not take such potentially harmful
actions. Instead, it is likely that people rely on action-oriented perceptions of affordances.

An affordance refers to the qualities (real or perceived) of the environment (or object).(14) These
qualities/properties determine how the environment/object can be used. For example, a chair of
sufficient size and stability affords sitting or climbing upon. Along similar lines, a gap in traffic
of sufficient distance may afford crossing of the roadway. Affordances are egocentric. This
means that a chair that affords sitting to a child may not afford sitting for an adult. Similarly, a
gap in traffic may afford crossing for an able-bodied young adult but not for an older adult with
mobility impairment. It is very important to note that although these qualities of the environment
are directly perceived, they may or may not be real. As a result, the affordances of an
environment can lead a person to take an incorrect or unsafe action. For example, imagine an
elongated door handle with the word “push” above it. A user will likely directly perceive that the
door handle appears to afford pulling and will attempt to pull the door open. This is an
“incorrect” action in the sense that the door must be pushed to open.

When they act upon perceived affordances, people generally produce a behavior that is adequate,
this is, unless the affordance leads the user to perform an unsafe action. According to this
perspective, behavior is goal-driven (e.g., I want to cross the street) rather than avoidance-driven
(e.g., I don’t want to be struck by that car). This results in generally adequate choices (e.g., I can
cross the roadway now). Of course, irregular outside influences have the potential to increase the

10
salience of avoidance-driven choices (e.g., seeing a police officer during a known target
jaywalking enforcement zone).

Perceived affordances are, of course, not the same as actual safety. People often incorrectly
perceive an action to be safe, when indeed, it is not. For example, as Tyrrell, Wood, and
Carberry found, at night, pedestrians frequently overestimate the distance at which drivers are
able to see them.(15) These perceptions can lead pedestrians to take unsafe actions. As such, it is
important to consider affordances when designing an environment. This includes affordances
that influence safety, aesthetics, and functionality. Take, for example, Norman’s classic teapot
example. Norman describes a teapot that has the handle and the spout on the same side of the
pot.(16) While the design is aesthetically pleasing, the pot is not functional, and could actually
pose quite a burn hazard. This example can easily be applied to the roadway and pedestrian
travel area. While it may be desirable to make pedestrian pathways pleasing to the eye, this
should not be the top priority when designing roadway environments. Instead, pathways should
be developed to increase the ability of pedestrians to safely (and quickly) move from an
originator to a destination. In reaching this goal, aesthetically pleasing environmental furniture
could be used. For example, attractive shrubbery might be placed along a curb to dissuade people
from crossing midblock (by reducing the roadway crossing affordance). Ultimately, roadway
environments should maximize the roadway-crossing safety while not inhibiting pedestrian or
vehicle travel.

If one is able to design an area that maximizes the affordances related to safe pedestrian crossing
behaviors (and minimizes those that increase more hazardous crossing), pedestrians should be
more likely to make safe crossings. Furthermore, this should reduce the need for hazard, or
warning signs. If a better understanding of pedestrian crossing affordances is gained,
modifications can be made to increase pedestrian safety, including the implementation of marked
crossings (in optimal locations that encourage use) or midblock crossing inhibitors (encouraging
pedestrians to cross at marked intersection crossings), while not inhibiting pedestrian travel.
Further, if pedestrian crossing schemas can be identified, possible educational interventions can
be developed.

An iterative process is presented here that explored the factors surrounding pedestrian crossings
at 20 different intersections in the Washington, DC, metropolitan area. Data were collected in
three separate phases. Each phase is discussed separately. A combined summary of each of the
phases is presented later. The influences of environmental factors on pedestrian crossing
locations are discussed. The individual relationships of these factors and where (i.e., intersection
or non-intersection) and how (e.g., walk, run, etc.) pedestrians cross the road are examined. In
addition, models are provided to predict where pedestrians are likely to cross given a set of
roadway environmental factors. The ability to readily predict problematic unmarked crossing
locations before crashes occur could be useful in the proactive deployment of mitigation
strategies. These may include educational interventions or engineering countermeasures, which
have been shown to be effective (e.g., Pelican, Puffin, and Zebra crossings, or refuge islands).(17)

11
12
 PEDESTRIAN CROSSING BEHAVIORS: PHASE 1

To assess pedestrian crossing behaviors, eight different locations in Washington, DC, were
selected. Pedestrian crossings were video recorded and later coded at each of the locations. The
following section describes this process as well as where pedestrians are likely to cross the
roadway at each of these locations. Later, these results are compared with more data that were
collected in phases 2 and 3.

DATA COLLECTION LOCATIONS

Eight different locations in Washington, DC, were selected as areas in which data would be
collected. Each area must have met several requirements to be included. These requirements
were the following:

 The area must be captured by a District of Columbia Department of Transportation

(DDOT) traffic management/closed-circuit television (CCTV) camera. This provided the
opportunity to record video feed directly from the cameras.

 Each camera must be positioned at an intersection that clearly captures a marked

intersection crosswalk in the foreground.

 Each camera must capture a second marked crosswalk. This ensured that all crossings
between two marked intersections could be captured.

 Each camera must capture enough traffic light information to determine walk/don’t walk
phases at the intersection crosswalk in the foreground.

Other roadway characteristics were not considered when selecting data collection areas. It is
possible that factors such as the number of lanes and directionality (one versus two lanes)
influence pedestrian crossing behaviors. As a result, these factors were not intentionally included
or excluded when selecting data collection areas. Rather, it was hoped that the influence of these
factors would present itself in the data.

Video feed was recorded from each of the eight cameras from 8:30 a.m. to 5:00 p.m. and from
8:00 p.m. to 11:00 p.m. Eastern Standard Time on February 9, 2012. Following this first day of
recording, all cameras recorded from 7:00 a.m. to 5:00 p.m. and from 8:00 p.m. to 11:00 p.m.
Eastern Standard Time, beginning February 10, 2012, and ending February 22, 2012.

The following section describes each of the eight data collection locations, which are labeled in
no particular order.

Location 1

The first data collection area is located at the intersection of 3rd Street Northeast and H Street
Northeast in Washington, DC. Figure 2 is an aerial view of this data collection area. Third Street
Northeast runs north/south and is located in the center of the image. H Street Northeast runs
east/west and is also located in the center of the image. The solid red rectangle highlights the
intersection of 3rd Street Northeast and H Street Northeast. The dotted red rectangle highlights

13
the intersection of 4th Street Northeast and H Street. The DDOT camera was positioned on the
southwest corner of the 3rd and H intersection facing east (see figure 3). As a result, pedestrians
making north/south crossings on H Street between 3rd Street and 4th Street were captured. The
distance from 3rd Street and 4th Street is approximately 352 ft. Figure 3 shows marked crossings
are present at both 3rd Street Northeast (solid red line) and 4th Street Northeast (dotted red line).
The AADT in this area is 15.8 (expressed in thousands and rounded to the nearest 100).(18)

Original image: ©2012 Google®; map annotations provided by SAIC.

Figure 2. Photo. Aerial view of Location 1.(19)

The neighborhood around this location contains both residential buildings and businesses. (It is
located only a few blocks from Union Station.) However, few businesses/offices along H Street
might attract pedestrian traffic. Along the entire block on north side of H Street, a large
construction project was underway throughout data collection. The result of this construction has
two relevant impacts on the present study. First, there are nearly no trip originators or
destinations on this side of the block for persons other than worksite employees. Second, a
barrier was placed between the sidewalk just outside the construction and the roadway. A break
in the barrier near the 3rd and H intersection allows pedestrians to enter and exit a nearby bus
stop. The southeast corner of this intersection has both a liquor store and a Capital Bikeshare™
station. (Capital Bike Share is a local subscription-based bicycle sharing program that allows
users to pick up and drop off bicycles at more than 140 different Washington, DC, metropolitan
area locations.) No other notable trip originators or destinations are located on the south side of
H Street.

14
 Figure 3. Photo. Still image captured from the Location 1 camera.

The northern portion of the relevant block of H Street Northeast (i.e., vehicles traveling from east
to west) provides two lanes of through traffic and an additional partial lane near the western end
of the block. This third lane functions primarily as an area for buses to stop to pick up
passengers. As vehicles continue to travel west (behind the view of the camera), the three lanes
continue and two additional bus stops are present. The southern portion of the relevant block of
H Street Northeast (i.e., vehicles traveling from west to east) provides two lanes of travel and a
third parking lane. Just west of the relevant intersection, two through lanes and a right turn only
lane are present. A bus stop is located just before the intersection.

The marked intersection of interest at 3rd and H Streets (as highlighted by the solid rectangular
box in figure 3 is 61 ft long (curb to curb). The walk phase is illuminated for 20 s, and the don’t
walk phase is illuminated for 69 s. Figure 4 illustrates what a pedestrian might see as he or she
attempts to cross from the north side of H Street Northeast to the south side of the street along
3rd Street Northeast. The image shows a pedestrian waiting to cross to the north side of the street
and looking for oncoming traffic.

15
 Figure 4. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side of H Street Northeast to the south side of the street along 3rd Street
Northeast.

Location 2

The second data collection area is located at the intersection of 7th Street Northwest and New
York Avenue Northwest in Washington, DC. Figure 5 is an aerial view of this data collection
area. Seventh Street Northwest runs north/south and is located in the center of the image. New
York Avenue runs northeast/southwest and is also located near the center of the image. The solid
red rectangle highlights the intersection of 7th Street Northwest and New York Avenue
Northwest. The dotted red rectangle highlights the intersection of 6th Street Northwest and New
York Avenue Northwest. The DDOT camera was positioned on the southwest corner of the 7th
and Mt. Vernon Place intersection facing east (see figure 6). (Note that because New York
Avenue is a “diagonal” street, it meets Mt. Vernon Place at Mount Vernon Square as can be seen
in figure 5.) As a result, pedestrians making north/south crossings on New York Avenue between
7th Street and 6th Street were captured. The distance between 7th Street and 6th Street is
approximately 530 ft. Figure 6 shows marked crossings are present at both 7th Street Northwest
(solid red rectangle) and 6th Street Northwest (dotted red rectangle). The AADT in this area is
24.2 (expressed in thousands and rounded to the nearest 100).(18)

16
Original image: ©2012 Google®; map annotations provided by SAIC.
Figure 5. Photo. Aerial view of Location 2.(20)

The neighborhood surrounding Location 2 contains both residential buildings and businesses.
The two largest attractions in the vicinity are Mount Vernon Square and the Walter E.
Washington Convention Center. Along the south side of New York Avenue are two smaller pay
parking lots. In addition, several older buildings (some vacant) are not likely to be substantial
trip originators or destinations. Along the north side of New York Avenue, several buildings
under construction are not likely to serve as trip originators or destinations for persons other than
worksite employees. There is also a smaller pay parking lot, a vehicle service station, vacant
buildings, and a business that functions primarily during evening hours. As a result, there are few
trip originators or destinations along this side of the roadway.

17
 Figure 6. Photo. Still image captured from the Location 2 camera.

The northern portion of the relevant block of New York Avenue Northwest (i.e., with vehicles
traveling from east to west) provides two lanes of through traffic. An additional third lane is
present that provides metered parking and a right turn only lane toward the western end of the
block. As vehicles continue to travel west (behind the view of the camera), the two travel lanes
continue. The southern portion of the relevant block of New York Avenue Northwest (i.e., with
vehicles traveling from west to east) provides two lanes of travel and a third parking lane. Just
west of the relevant intersection, two through lanes are present. There is also a small “soft”
median (i.e., a small area in the roadway that is not raised/protected by concrete) between the
eastbound and westbound travel lanes of New York Avenue. Also of relevance, the rightmost
lane traveling northbound on 7th Street Northwest is right turn only. That is, there is a dedicated
lane of traffic turning onto New York Avenue.

The marked intersection of 7th Street and New York Avenue of interest (as highlighted by the
solid rectangular box in figure 6) is 81 ft 3 inches long (curb to curb). The walk phase is
illuminated for 46 s, and the don’t walk phase is illuminated for 55 s. Figure 7 illustrates what a
pedestrian might see as he or she attempts to cross from the north side of New York Avenue
Northwest to the south side of the street along 7th Street Northwest. The traffic traveling
northbound (closest to the crosswalk) is waiting for the pedestrians to clear before completing a
right turn.

18
 Figure 7. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side of New York Avenue Northwest to the south side of the street along 7th
Street Northwest.

Location 3

The third data collection area is located at the intersection of 7th Street Northwest and Rhode
Island Avenue Northwest in Washington, DC. An aerial view of this data collection area can be
seen in figure 8. Seventh Street Northwest runs north/south and is located in the center of the
image. New York Avenue runs northeast/southwest and is also located near the center of the
image. The solid red rectangle highlights the intersection of 7th Street Northwest and Rhode
Island Avenue Northwest. The dotted red rectangle highlights the intersection of 6th Street
Northwest and Rhode Island Avenue Northwest. The DDOT camera was positioned on the
northwest corner of the 7th and Rhode Island intersection facing east (see figure 9). As a result,
pedestrians making north/south crossings on Rhode Island Avenue between 7th and 6th Streets
were captured. The distance between 7th Street and 6th Street is approximately 550 ft. Figure 9
shows marked crossings are present at both 7th Street Northwest (solid red rectangle) and
6th Street Northwest (dotted red rectangle). The AADT in this area is 26.3 (expressed in
thousands and rounded to the nearest 100).(18)

In addition, Marion Street Northwest runs north/south between 6th and 7th Streets. Marion Street
is located to the south of Rhode Island Avenue. Northbound Marion Street traffic must turn

19
(either right or left) when it meets Rhode Island. This intersection does not contain a marked
pedestrian crossing. The distance between 7th Street and Marion Street is approximately 277 ft.

Original image: ©2012 Google®; map annotations provided by SAIC.

Figure 8. Photo. Aerial view of Location 3. Northwest between 7th and 6th Streets
Northwest.(21)

The neighborhood surrounding Location 3 consists of mostly residential buildings and a few
small businesses. Along the south side of Rhode Island Avenue, west of Marion Street, a
residential building complex fills the entire block. The south side of Rhode Island Avenue (east
of Marion Street) comprises a small grassy area. Along the north side of Rhode Island Avenue,
to the west of Marion Street, is a vacant lot. Directly north of the Rhode Island Avenue and
Marion Street intersection is construction of a new housing complex. A barrier was placed
between the sidewalk and the roadway in front of this construction area. However, there are
notable gaps in the barrier blocks. The remaining portion of the northeastern section of Rhode
Island Avenue consists of houses. Other trip originators or destinations include a convenience
store on the southwestern corner of the intersection of 7th Street Northwest and Rhode Island
Northwest. A public library is also located directly west of the intersection. In addition, a Metro
station (Shaw-Howard) is located on the northwestern adjacent block (behind the camera).

20
 Figure 9. Photo. Still image captured from the Location 3 camera.

The northern portion of the relevant block of Rhode Island Avenue Northwest (i.e., vehicles
traveling from east to west) provides three lanes of through traffic. An additional third left turn
only lane is present just east of the intersection, and there is a split to the right so that drivers can
travel westbound on R Street Northwest. As vehicles continue to travel west (behind the view of
the camera), the three travel lanes continue. The southern portion of the relevant block of New
York Avenue Northwest (i.e., vehicles traveling from west to east) provides three lanes of travel
continued from the roadway west of the intersection. A bus stop is located just to the east of the
intersection. A concrete median is present between the eastbound and westbound lanes of traffic.
This median is also present to the east of Marion Street Northwest.

The marked intersection of 7th Street and Rhode Island Avenue of interest (as highlighted by the
solid rectangular box in figure 9) is 48 ft 3 inches long (median refuge to curb). The walk phase
is illuminated for 46 s, and the don’t walk phase is illuminated for 71 s. The median between the
eastbound and westbound traffic is 7 ft wide. Figure 10 illustrates what a pedestrian might see as
he or she attempts to cross from the north side of Rhode Island Avenue Northwest to the south
side of the street along 7th Street Northwest.

21
Figure 10. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side of Rhode Island Avenue Northwest to the south side of the street along
7th Street Northwest.

Location 4

The fourth data collection area is located at the intersection of 13th Street Northwest and I (eye)
Street Northwest in Washington, DC. An aerial view of this data collection area can be seen
in Figure 11. Thirteenth Street Northwest runs north/south and is located toward the right of the
image. Traffic travels east/west on I Street Northwest and is located near the center of the image.
The solid red rectangle highlights the intersection of 13th Street Northwest and I Street
Northwest. The dotted red rectangle highlights the intersection of 14th Street Northwest and
I Street Northwest. The DDOT camera was positioned on the northeast corner of the 13th and I
intersection facing west (see figure 11). As a result, pedestrians making north/south crossings on
I Street between 13th and 14th Streets were captured. The distance between 13th Street and
14th Street is approximately 565 ft. Figure 12 shows marked crossings are present at both
13th Street Northwest (solid red line) and 14th Street Northwest (dotted red line). The AADT in
this area is 11.3 (expressed in thousands and rounded to the nearest 100).(18)

22
Original image: ©2012 Google®; map annotations provided by SAIC.
Figure 11. Photo. Aerial view of Location 4.(22)

The neighborhood surrounding Location 4 consists of mostly businesses. Along the south side of
I Street Northwest, there are two large office buildings. They contain a variety of businesses,
including restaurants, automatic teller machines (ATM), and a staffing solutions/temporary
employment center. The entire north side of the relevant I Street block is bordered by Franklin
Park.

Figure 12. Photo. Still image captured from the Location 4 camera.

I Street Northwest is a one-way street with traffic flowing from east to west. To the east of the
relevant 13th Street intersection, vehicles travel in two through lanes along I Street. Parking is
available on both the north and south sides of the roadway. As traffic moves west to the data

23
collection zone between 13th and 14th Streets, a third travel lane becomes available. Parking is
also available on the south side of I street. A bus lane with multiple stops is present on the north
side of I Street. The bus lane continues to the block west of 14th Street as parking space.

The marked intersection of 13th Street and I Street of interest (as highlighted by the solid
rectangular box in figure 12 is 54 ft 5 inches long (curb to curb). The walk phase is illuminated
for 32 s, and the don’t walk phase is illuminated for 45 s. Figure 13 shows the view of the
crossing of I Street Northwest at 13th Street Northwest from the southeast corner of the
intersection.

Figure 13. Photo. Image of the westernmost crossing of I Street Northwest at 13th Street
Northwest. The photo is taken from the southwestern corner of the intersection.

Location 5

The fifth data collection area is located at 14th Street Northwest and H Street Northwest in
Washington, DC. An aerial view of this data collection area can be seen in figure 14. Fourteenth
Street Northwest runs north/south and is located toward the right of the image. H Street
Northwest travels east/west and is located near the center of the image. The solid red rectangle
highlights the intersection of 14th Street Northwest and H Street Northwest. The dotted red
rectangle highlights the intersection of 15th Street Northwest and H Street. The DDOT camera

24
was positioned on the southeast corner of the 14th and H intersection facing west (see
figure 15). As a result, pedestrians making north/south crossings on H Street between 14th and
15th Streets were captured. The distance between 14th Street and 15th Street is approximately
391 ft. Figure 15 shows marked crossings are present at both 14th Street Northwest (solid red
line) and 15th Street Northwest (dotted red line). The AADT in this area is 15.1 (expressed in
thousands and rounded to the nearest 100).(18)

Original image: ©2012 Google®; map annotations provided by SAIC.

Figure 14. Photo. Aerial view of Location 5.(23)

The neighborhood surrounding Location 5 consists of mostly businesses. Along the south side of
H Street Northwest, there are several large buildings. These buildings contain a variety of
businesses, including restaurants, offices, and a public parking garage. The north side of H Street
is also made up of office buildings. These buildings contain the city center offices, an engraving
shop, a men’s clothing shop, a hair salon, a mobile telephone store, and a parking garage.

25
 Figure 15. Photo. Still image captured from the Location 5 camera.

H Street Northwest is a one-way street with traffic flowing from west to east. H Street between
15th and 14th Streets contains three travel lanes and parking both to the north and south. As
vehicles approach the 14th Street intersection, the parking along the north side of H Street
becomes a left turn only lane, and the parking along the south side of H Street becomes a right
turn only lane. As traffic continues past this intersection traveling east, there are three travel
lanes and a parking lane on the north side of H Street.

The marked intersection of 14th Street and H Street of interest (as highlighted by the solid
rectangular box in figure 15) is 61 ft long (curb to curb). The walk phase is illuminated for 53 s,
and the don’t walk phase is illuminated for 50 s.

Figure 16 illustrates what a pedestrian might see as he or she attempts to cross from the south
side of H Street Northwest to the north side of the street along 14th Street Northwest.

26
Figure 16. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the south side of H Street Northwest to the north side of the street along 14th Street
Northwest.

Location 6

The sixth data collection area is located at 14th Street Northwest and New York Avenue
Northwest in Washington, DC. An aerial view of this data collection area can be seen in
figure 17. Fourteenth Street Northwest runs north/south and is located toward the center of the
image. New York Avenue Northwest travels northeast/southwest and is located near the center/
top of the image. The solid red rectangle highlights the intersection of 14th Street Northwest and
New York Avenue Northwest. The dotted red rectangle highlights the intersection of 14th Street
Northwest and G Street Northwest. The DDOT camera was positioned on the northeast corner of
the 14th and New York intersection facing south (see figure 18). As a result, pedestrians making

27
east/west crossings on 14th Street between New York Avenue and G Street were captured. The
distance between New York Avenue and G Street is approximately 294 ft. Figure 18 shows
marked crossings are present at both New York Avenue Northwest (solid red line) and G Street
Northwest (dotted red line) along 14th Street Northwest. The AADT was not available for this
specific block. However, the AADT for 14th Street Northwest on the block north of New York
Avenue is 27.2 (expressed in thousands and rounded to the nearest 100).(18) In addition, the
AADT for 14th Street Northwest two blocks south of New York Avenue is 35.7 (expressed in
thousands and rounded to the nearest 100).(18) This suggests that the AADT for this block is in
the middle 30s range. (An AADT of 31.5 was used for analysis purposes.)

Original image: ©2012 Google®; map annotations provided by SAIC.

Figure 17. Photo. Aerial view of Location 6.(24)

The neighborhood surrounding Location 6 consists of mostly businesses. The west side of
14th Street contains a restaurant serving primarily sandwiches, a mobile telephone repair shop, a
fitness center, a parking garage, and office space. The east side of 14th Street contains a major
pharmacy chain store, a bank, a shoe store, a café/restaurant, and office space.

28
 Figure 18. Photo. Still image captured from the Location 6 camera.

Fourteenth Street Northwest is a two-way street with traffic flowing north/south. There are three
northbound travel lanes. A bus stop is located in the far right northbound lane near the
intersection of 14th and New York. There are three southbound travel lanes, and an additional
left turn lane that spans the entire block. A bus stop is located in the far right southbound lane
near the intersection of 14th and G Streets.

The marked intersection of 14th Street and New York Avenue of interest (as highlighted by the
solid rectangular box in figure 18) is 83 ft long (curb to curb). The walk phase is illuminated for
31 s, and the don’t walk phase is illuminated for 56 s. Figure 19 shows the southernmost crossing
of 14th Street Northwest at New York Avenue Northwest.

29
 Figure 19. Photo. An image of the southernmost crossing of 14th Street Northwest at New
York Avenue Northwest. The photo is taken from the northwestern corner of the
intersection.

Location 7

The seventh data collection area is located at 19th Street Northwest and Pennsylvania Avenue
Northwest in Washington, DC. An aerial view of this data collection area can be seen
in figure 20. Nineteenth Street Northwest runs north/south and is located toward the center of the
image. Pennsylvania Avenue Northwest runs northwest/southeast and is located near the
center/top of the image. The solid red rectangle highlights the intersection of 19th Street
Northwest and Pennsylvania Avenue Northwest. The dotted red rectangle highlights the
intersection of 19th Street Northwest and H Street Northwest. The DDOT camera was positioned
on the northwest corner of the 19th and Pennsylvania intersection facing south (see figure 21).
As a result, pedestrians making east/west crossings on 19th Street between Pennsylvania Avenue
and H Street were captured. The distance between Pennsylvania Avenue and H Street is
approximately 145 ft. Figure 21 shows marked crossings are present at both Pennsylvania
Avenue Northwest (solid red line) and H Street Northwest (dotted red line) along 19th Street
Northwest. Unlike the first six collection sites, the far marked crosswalk is not controlled by a
signal (nor is it controlled by a stop sign). The AADT in this area is 11.9 (expressed in thousands
and rounded to the nearest 100).(18)

30
Original image: ©2012 Google®; map annotations provided by SAIC.
Figure 20. Photo. Aerial view of Location 7.(25)

The neighborhood surrounding Location 7 consists of mostly businesses. The west side of
19th Street contains a single, large office building. In addition, several large flower planters and
seating areas prevent easy access to the roadway from the sidewalk (although it is not impossible
to access the roadway). The east side of 19th Street contains a small park area that is blocked
from the sidewalk access along 19th Street by shrubbery.

Figure 21. Photo. Still image captured from the Location 7 camera.

Nineteenth Street Northwest is a one-way street with traffic flowing south. There are three travel
lanes and a single parking lane on the east side of the roadway. A bus stop is present on the

31
western portion of the roadway. The roadway north of Pennsylvania (behind the camera view)
contains two travel lanes and parking on both sides of the roadway.

The marked intersection of 19th Street and Pennsylvania Avenue of interest (as highlighted by
the solid rectangular box in figure 21) is 43 ft long (curb to curb). The walk phase is illuminated
for 44 s, and the don’t walk phase is illuminated for 52 s. Figure 22 illustrates what a pedestrian
might see as he or she attempts to cross from the west side of 19th Street Northwest to the east
side of the street along Pennsylvania Avenue Northwest.

Figure 22. Photo. Illustration of what a pedestrian might see as he or she attempts to
crossing from the west side of 19th Street Northwest to the east side of the street along
Pennsylvania Avenue Northwest.

Location 8

The eighth data collection area is located at Connecticut Avenue Northwest and Oliver Street
Northwest in Washington, DC. An aerial view of this data collection area can be seen
in figure 23. Connecticut Avenue Northwest runs north/south and is located toward the center of
the image. Oliver Street Northwest travels east/west and is located toward the top of the image.
The solid red rectangle highlights the intersection of Oliver Street Northwest and Connecticut
Avenue Northwest. The dotted red rectangle highlights the intersection of Northampton Street
Northwest and Connecticut Northwest. The DDOT camera was positioned north of the

32
Connecticut and Oliver intersection facing south (see figure 24). As a result, pedestrians making
east/west crossings on Connecticut Avenue between Oliver Street and Northampton Street
Northwest were captured. The distance between Oliver Street and Northampton Street is
approximately 433 ft. Figure 24 shows marked crossings are present at both Oliver Street
Northwest (solid red line) and Northampton Street Northwest (dotted red line) along Connecticut
Avenue Northwest. Unlike the first six data collection sites, the far marked crosswalk is not
controlled by a signal (nor is it controlled by a stop sign). The far marked crossing uses a flag
crossing system (see figure 25). The AADT in this area is 24.2 (expressed in thousands and
rounded to the nearest 100).(18)

Original image: ©2012 Google®; map annotations provided by SAIC.

Figure 23. Photo. Aerial view of Location 8.(25)

The neighborhood surrounding Location 8 consists of a mix of businesses and homes. The west
side of Connecticut Avenue contains banks, small restaurants, and other small businesses. The
east side of 19th Street contains a bank. This area has a more suburban feel than the other seven
urban environments in this phase.

33
 Figure 24. Photo. Still image captured from the Location 8 camera.

Figure 25. Photo. An image of the (far) flag crossing at Connecticut Avenue Northwest and
Northampton Street Northwest.

34
Connecticut Avenue Northwest is a two-way street flowing north/south. There are three
northbound travel lanes and three southbound travel lanes. The marked intersection of Oliver
Street and Connecticut Avenue of interest (as highlighted by the solid rectangular box in
figure 24) is 70 ft long (curb to curb). The walk phase is illuminated for 19 s, and the don’t walk
phase is illuminated for 79 s. Figure 26 illustrates what a pedestrian might see as he or she
attempts to cross from the west side of Connecticut Avenue Northwest to the east side of the
street along Oliver Street Northwest.

Figure 26. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the west side of Connecticut Avenue Northwest to the east side of the street along
Oliver Street Northwest.

DATA COLLECTION VALIDATION

Video recording pedestrian crossings is not a common method for assessing pedestrian
behaviors. Further, to the authors’ knowledge, no other studies have attempted to use existing
traffic cameras/CCTV to assess pedestrian behavior. As such, it was deemed important to assess
the validity of camera footage as a data collection methodology.

35
Researchers manually scored pedestrian crossing behaviors at Location 4 (13th Street and
I Street Northwest) and Location 5 (14th Street and H Street Northwest) onsite. These two
locations were selected for their physical proximity to one another. Researchers recorded the
pedestrians’ crossings and their interactions with vehicles (described in more detail in the next
section) over three 15-min periods at both locations. The in-vivo recordings were made the
morning of February 20, 2012 (President’s Day). Unfortunately the camera recording Location 5
was facing the incorrect direction on this day. As a result, there is no video data to compare with
the live, onsite scoring. However, the onsite pedestrian crossing scoring was compared with the
DDOT video at Location 4.

At Location 4, there was a 100-percent agreement in the classification of the pedestrian crossings
for both the first and last 15-min segments (18 and 21 crossings, respectively). However, in the
second 15-min session, there was a single discrepancy; the onsite coding resulted in five total
pedestrian crossings, and the video coding resulted in four crossings. The categorization of the
four video crossings was the same as those in the in-vivo coding. This 98-percent agreement
between the two coding methodologies provides evidence that video coding is a reliable
methodology for coding and characterizing pedestrian crossing behaviors.

VIDEO DATA CODING

Many different types of pedestrian crossings and pedestrian interactions with vehicles in, and
along, the roadway can be recorded and classified. Although it is difficult to code pedestrian
crossings in an exhaustive manner, the current study sought to record enough information to
interpret general crossing behaviors. For each pedestrian crossing, multiple factors were
recorded. The following subsections describe each of these factors and how data were coded.

Crossing Factors

Data were coded between the marked intersection closest to the DDOT camera and a far marked
intersection (as denoted in each of the location descriptions). Pedestrian crossings were not
counted in the far marked intersections. Only completed crossings were recorded and included in
data analysis. In other words, if a person walked partially into the roadway to hail a cab, it was
not included as a pedestrian crossing.

Location/Crossing Area

The location where pedestrians crossed the road was classified in one of three categories:

 Marked Intersection. This included any crossing that occurred in any portion of the
marked crosswalk at the intersection closest to the DDOT camera. Crossings within one
car length of the marked crosswalk were also included in this category. This is simply
because some of the marked intersections contained enough pedestrian traffic that all
persons could not comfortably cross within the marked crosswalk.

 Unmarked Intersection. This was any crossing at an intersection without a marked

crosswalk. This type of crossing only applied to Location 3.

36
  Unmarked Non-Intersection. This type of crossing refers to any pedestrian crossing that
occurred at a non-intersection location without a marked crossing.

Traffic

The status of vehicular traffic was also noted.

 With Traffic. This is a crossing that does not conflict with traffic flow. An example is a
crossing in the crosswalk with the walk sign activated.

 Against Traffic. This is a crossing that conflicts with traffic flow. An example is
crossing in the crosswalk while oncoming traffic has a green light and the don’t walk sign
is activated. (The don’t walk sign must be in the steady state to be counted as in the don’t
walk phase.)

 Traffic Flow Change. This is a crossing that starts either with or against traffic, but ends
with the opposite traffic flow. An example of this type of crossing is one in which a
pedestrian starts crossing the street with the don’t walk sign activated and completes the
crossing with the walk sign activated.

Yielding Behavior

Yielding behavior of both pedestrians and vehicles was recorded. A note on where the yielding
occurred and the light status was recorded. Vehicle or pedestrian right of way can be established
using this information. (Note that right-of-way laws vary by State. However, in the present
context, the legality of the crossing is not considered.)

 Pedestrian Yielding. This was recorded when a pedestrian yielded to a vehicle. An

example is a pedestrian pausing in the roadway to allow a turning vehicle to complete a
turn, prior to the pedestrian completing a crossing.

 Vehicle Yielding. This was recorded when a vehicle yielded to a pedestrian. An example
is a vehicle pausing mid-turn to allow a pedestrian to complete the crossing.

Evasive Pedestrian Actions

Actions made by pedestrians to avoid conflict or potential collisions were also recorded. A
distinction between evasive actions in the first half and second half of the crossing was made.

 Running/Accelerated Walking. This was recorded when a pedestrian either ran or

noticeably increased walking speed to avoid collision or potential conflict with a vehicle.
Note that pedestrians who ran across the intersection, but were not taking evasive actions
(e.g., joggers), were not coded as taking this type of evasive action.

 Abrupt Stopping. This was recorded when a pedestrian stopped abruptly in the roadway
to avoid collision or potential conflict with a vehicle.

37
  Directional Change. This was recorded when a pedestrian changed travel direction, once
in the roadway, to avoid collision or potential conflict with a vehicle. An example of this
is a pedestrian stepping into the roadway and then subsequently returning to the curb
(direction change) after noting an oncoming vehicle.

Evasive Vehicle Actions

Evasive actions taken by vehicles were also recorded. A distinction between evasive actions
taken by the vehicle closest to the pedestrian and a following vehicle (if any) was made.

 Abrupt Braking (First Vehicle). This was recorded when a vehicle was forced to
abruptly brake (not a gradual deceleration) to avoid collision with a pedestrian.

 Abrupt Braking (Second Vehicle). This was recorded when the action of the vehicle
closest to the pedestrian required the next vehicle to brake abruptly to avoid collision.

 Directional Change (First Vehicle). This was recorded when a vehicle swerved (or
otherwise modified direction) to avoid collision with a pedestrian.

 Directional Change (Second Vehicle). This was recorded when the action of the vehicle
closest to the pedestrian required the next vehicle to modify direction to avoid collision.

Other Relevant Information

Other relevant information related to the pedestrian crossings was also recorded. For example, if
an ambulance or presidential motorcade interrupted regular traffic flow, this was marked to
accommodate potential irregular data coding.

Dates/Times Coded

Only sub-portions of the vast quantity of video recorded data were coded, owing to both time
and project scope requirements of the current study. The scope of the present study included only
daytime pedestrian crossings. As such, nighttime data were not examined.

Each of the eight cameras was coded for the entire day of February 9, 2012 (daytime hours only).
The peak pedestrian traffic times were selected based on this information. As a result of
pedestrian traffic and sunrise/sunset times, data were coded from 7:30 to 9:30 a.m., 11:30 to
1:30 p.m., and 4:00 to 5:00 p.m. Locations 1, 3, 4, 6, 7, and 8 were coded Monday through
Friday, February 9 through February 21, 2012. Location 2 was not coded on February 10 or 13
because of a camera malfunction that resulted in a “frozen” image. In other words, no new feed
from the camera was recorded. Location 5 was not coded after February 13 because the camera
was facing a different direction that did not allow for data coding based on the previously
mentioned location requirements.

Weather

Table 1 describes the weather and corresponding sunrise/sunset times for each of the days that
data were coded.

38
 Table 1. Relevant sunrise, sunset, and weather for each of the coded data collection days.
High Low Precipitation
Sunrise Sunset Temperature Temperature General Accumulation
Date (a.m.) (p.m.) (ᵒF) (ᵒF) Weather (inches)
Partly
2/09/2012 7:07 5:39 47 33 —
Cloudy
2/10/2012 7:05 5:40 47 32 Rain 0.11
Partly
2/13/2012 7:02 5:43 49 24 —
Cloudy
Mostly
2/14/2012 7:01 5:44 56 38 —
Cloudy
2/15/2012 7:00 5:45 54 39 Windy —
2/16/2012 6:58 5:46 47 42 Rain .09
2/17/2012 6:57 5:48 54 39 Fog —
Partly
2/20/2012 6:53 5:51 50 34 —
Cloudy
2/21/2012 6:52 5:52 50 33 Windy —
— Indicates no recorded precipitation.

Inter-Rater Reliability

Three different researchers coded the video data. Each of the researchers had been to the actual
data collection/DDOT camera locations. To ensure that each rater was coding instances
similarly, inter-rater reliability was assessed. Three randomly selected 5-min segments of video
were coded by all three researchers. There was a 100-percent agreement among the raters. This is
not to say that all three raters agreed on every crossing across all of the data. However, it does
provide evidence that all three raters were using the same crossing factor definitions, and the
video provided enough information to clearly identify the circumstances of the pedestrian
crossings.

Each of the raters coded approximately one-third of the total video data in this phase.

RESULTS

The follow subsections describe the overall crossing behavior results from phase 1.

Descriptive Data

Table 2 summarizes the crossing behaviors recorded during the coded data collection times. All
values are combined over all eight data collection locations. The following subsections explain
the basic descriptive data for each of the eight locations in further detail.

39
 Table 2. Summary of pedestrian crossings combined across all eight data collection locations.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 51,956 119 9,272 855 13 18 4 — 2 4
+traffic flow
527 10 42 128 1 1 — — — —
change
Don’t Walk 5,480 4 15 334 68 41 1 — — —
+traffic flow
1,041 17 7 27 13 1 — — — —
change
Sum 59,004 150 9,336 1,344 95 61 5 — 2 4
Unmarked Non-Intersection
With Traffic 2,924 16 16 109 5 5 1 — — —
+traffic flow
120 4 2 18 — — — — — —
change
Against Traffic 1,126 35 14 129 10 3 — — — —
+traffic flow
229 17 — 11 6 — — — — —
change
Sum 4,399 72 32 267 21 8 1 — — —
Unmarked Intersection
With Traffic 954 3 1 9 — — — — — —
+traffic flow
84 3 3 5 — 1 — — — —
change
Against Traffic 1,251 67 13 47 1 3 — — — —
+traffic flow
33 3 — — — — — — — —
change
Sum 2,322 76 17 61 1 4 — — — —
Grand Total 65,725 298 9,385 1,672 117 73 6 — 2 4
— Indicates no actions of this type were recorded.

40
Location 1

Total Crossings:

There were a total of 1,110 pedestrian crossings on H Street Northeast between 3rd and
4th Streets Northeast during the coded data collection times (a total of just over 48 h of video).
(See table 3 for a summary of the crossings.) Of these crossings, 976 (or 87.93 percent) occurred
within the marked intersection (i.e., marked crosswalk). These crossings can be broken down
into when they occurred during the light phases. During the walk phase, 834 pedestrians crossed
the marked intersection. An additional 94 crossings occurred in the marked intersection during
the don’t walk phase. In addition, 48 crossings occurred in the marked intersection that included
a traffic flow change. Of these crossings, eight began in the walk phase and ended in the don’t
walk phase. An additional 40 began in the don’t walk phase and concluded in the walk phase. In
other words, pedestrians who cross during a crosswalk light phase change were approximately
equally likely to start the crossing during the don’t walk phase as during the walk phase.

At Location 1, 134 (or 12.07 percent) of the total 1,110 crossings took place at the unmarked
non-intersection areas. Of these crossings, 58 took place with traffic, and 64 took place against
traffic. A portion of the pedestrian crossings that took place at an unmarked non-intersection area
involved a traffic flow change. Five of these crossings began with traffic and ended against
traffic, and seven began against traffic and ended with traffic.

Yielding:

In total, there were 45 yielding behaviors (combined across pedestrians and vehicles). In
34 instances, a vehicle yielded to a pedestrian crossing in the marked intersection during a walk
phase. In other words, vehicles were appropriately yielding to the pedestrians crossing during the
appropriate light phase. There was a single case where a pedestrian yielded to a vehicle while
crossing in the marked intersection during a walk phase.

Ten pedestrian crossings with yielding occurred outside the marked intersection. In four cases,
pedestrians yielded to vehicles in the unmarked non-intersection. In six instances, vehicles
yielded to pedestrians who were crossing in the unmarked non-intersection.

Evasive Pedestrian Actions:

A total of 97 pedestrians took evasive actions at Location 1. (Note that some pedestrians took
more than one evasive action, e.g., returning to the curb followed by running.) There were
29 instances of pedestrian running/accelerated walking while in the marked intersection during
the walk phase. Although a pedestrian taking an evasive action while crossing the street during a
walk phase in the marked intersection may seem counterintuitive, there were many instances of
“courtesy” acceleration. In other words, it appeared that pedestrians would run, or accelerate,
through a crossing to allow turning vehicles to complete their turn during the signal phase. There
were 20 instances of pedestrian running/accelerated walking while in the marked intersection
during the don’t walk phase. In six instances, pedestrians ran/accelerated while in the marked
intersection and a traffic flow change was involved. Of these, one began in the walk phase and
was completed during the don’t walk phase, and five began in the don’t walk phase and were

41
completed in the walk phase. In addition, there were 28 instances of running/accelerated walking
while in an unmarked non-intersection area.

There were also eight instances where pedestrians abruptly stopped while crossing the roadway.
In a single instance, a pedestrian abruptly stopped while walking in the marked intersection
during the walk signal phase to yield to a vehicle. In two instances, pedestrians abruptly stopped
while walking in the marked intersection during a traffic flow change (from the don’t walk phase
to the walk phase). An additional four pedestrians abruptly stopped in the marked intersection
during the don’t walk phase. Finally, four pedestrians stopped during crossings at unmarked non-
intersection locations.

In five instances, pedestrians made a travel directional change (e.g., returned to the curb). One of
these occurred at an unmarked non-intersection area, three occurred at the marked intersection
during the don’t walk phase, and one occurred at the marked intersection during the walk phase.

Evasive Vehicle Actions:

In a total of two instances, a vehicle took an evasive action to avoid a potential collision with a
pedestrian at Location 1. In one instance, a vehicle abruptly braked for a pedestrian crossing at
an unmarked non-intersection area. In the second instance, a vehicle abruptly braked for a
pedestrian crossing marked intersection during the don’t walk phase.

42
 Table 3. Summary of pedestrian crossings at Location 1.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 834 1 34 29 1 1 — — — —
+traffic flow
8 — — 1 — — — — — —
change
Don’t Walk 94 — — 20 4 3 1 — — —
+traffic flow
40 — — 5 2 — — — — —
change
Sum 976 1 34 55 7 4 1 — — —
Unmarked Non-Intersection
With Traffic 58 2 3 11 — — 1 — — —
+traffic flow
5 1 — — — — — — — —
change
Against Traffic 64 — 3 16 3 1 — — — —
+traffic flow
7 1 — 1 1 — — — — —
change
Sum 134 4 6 28 4 1 1 — — —
Location 1 Total 1,110 5 40 83 11 5 2 — — —
— Indicates no actions of this type were recorded.

43
Location 2

Total Crossings:

There were a total of 4,631 pedestrian crossings on New York Avenue Northwest between
7th Street Northwest and 6th Street Northwest during the coded data collection times. (See
table 4 for a summary of the crossings.) Location 2 was not coded on February 10 or 13 because
of a camera malfunction that resulted in a “frozen” image. In other words, no new feed from the
camera was recorded during this time frame. Of the recorded and coded crossings, 4,228 (or
91.3 percent) occurred within the marked intersection (i.e., marked crosswalk). These crossings
can be broken down into when they occurred during the light phases. During the walk phase,
4,071 pedestrians crossed entirely within the near marked intersection. An additional
81 crossings occurred entirely in the marked intersection during the don’t walk phase. In
addition, 76 crossings occurred in the marked intersection and included a traffic flow change. Of
these crossings, 52 began in the walk phase and ended in the don’t walk phase. An additional 24
began in the don’t walk phase and concluded in the walk phase. In other words, pedestrians who
crossed during a crosswalk light phase change were more likely to start the crossing during the
walk phase than the don’t walk phase.

At Location 2, 403 (or 8.7 percent) of the total 4,631 crossings took place at unmarked non-
intersection areas. Of these crossings, 289 took place entirely with traffic, and 74 took place
entirely against traffic. A portion of the pedestrian crossings that took place at an unmarked non-
intersection area involved a traffic flow change. Fourteen of these crossings began with traffic
and ended against traffic, and 26 began against traffic and ended with traffic.

Yielding:

In total, there were 342 yielding behaviors (combined across pedestrians and vehicles) at
Location 2. In 294 instances, a vehicle yielded to a pedestrian in the marked intersection during a
walk phase. In five instances, a pedestrian yielded to a vehicle while crossing in the marked
intersection during a walk phase. In addition, seven pedestrian crossings took place in the
marked intersection area and involved a traffic flow change. Of these crossings, one began in the
walk phase and ended in the don’t walk phase. An additional six began in the don’t walk phase
and concluded in the walk phase.

There were 25 instances of a pedestrian yielding to a vehicle while crossing at an unmarked non-
intersection location.

Evasive Pedestrian Actions:

A total of 268 pedestrians took evasive actions at Location 2. There were 198 instances of
pedestrian running/accelerated walking while in the marked intersection during the walk phase.
There were also 13 instances of pedestrian running/accelerated walking while in the marked
intersection during the don’t walk phase. In addition, 14 pedestrians crossed during a traffic flow
change in the marked intersection. Two people began crossing during the don’t walk phase and
completed during the walk phase, and 12 began crossing during the walk phase and completed
the crossing during the don’t walk phase. In addition, 25 pedestrians ran/accelerated at an
unmarked non-intersection area.

44
In addition, in 16 cases, pedestrians abruptly stopped while crossing the roadway. In two
instances, pedestrians abruptly stopped while walking in the marked intersection during the walk
signal phase to yield to a vehicle. Three pedestrians abruptly stopped in the marked intersection
during the don’t walk phase. There were also three cases of abruptly stopping during a traffic
flow change, one during the walk phase and completed the crossing during the don’t walk phase,
and two that began crossing during the don’t walk phase and completed during the walk phase.
Eight pedestrians abruptly stopped while crossing in an unmarked non-intersection location.

In six instances, pedestrians made a travel directional change (i.e., returned to the curb). Two of
these occurred in the marked intersection during the walk phase, and the remaining four occurred
at an unmarked non-intersection area.

Evasive Vehicle Actions:

In a total of 10 instances, a vehicle took an evasive action to avoid a potential collision with a
pedestrian at Location 2. All 10 of these instances occurred while the pedestrians were crossing
at the marked intersection during the walk phase.

45
 Table 4. Summary of pedestrian crossings at Location 2.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional
Running/ Braking Braking Change Directional
Total Accelerated Abrupt Directional (1st (2nd (1st Change
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) (2nd Vehicle)
Intersection
Walk 4,071 5 294 198 2 2 4 — 2 4
+traffic flow
52 1 — 12 1 — — — — —
change
Don’t Walk 81 — — 13 3 — — — — —
+traffic flow
24 6 — 2 2 — — — — —
change
Sum 4,228 12 294 225 8 2 4 — 2 4
Unmarked Non-Intersection
With Traffic 289 6 — 8 3 3 — — — —
+traffic flow
14 1 — — — — — — — —
change
Against Traffic 74 9 — 13 2 1 — — — —
+traffic flow
26 9 — 4 3 — — — — —
change
Sum 403 25 — 25 8 4 — — — —
Location 2 Total 4,631 37 294 250 16 6 4 — 2 4
— Indicates that no actions of this type were recorded.

46
Location 3

Total Crossings:

There were a total of 5,200 pedestrian crossings on Rhode Island Avenue Northwest between
7th Street Northwest and 6th Street Northwest during the coded data collection times. (See table
5 for a summary of the crossings.) Of these crossings, 2,646 (or 50.9 percent) occurred within the
marked intersection (i.e., marked crosswalk). These crossings can be broken down into where
they occurred during the light phases. During the walk phase, 2,317 pedestrians crossed within
the near marked intersection. An additional 298 crossings occurred in the marked intersection
during the don’t walk phase. In addition, 31 crossings occurred in the marked intersection that
included a traffic flow change. Of these crossings, 20 began in the walk phase and ended in the
don’t walk phase. An additional 11 began in the don’t walk phase and concluded in the walk
phase. In other words, pedestrians who crossed during a crosswalk light phase change were more
likely to start the crossing during the walk phase than the don’t walk phase.

At Location 3, 232 (or 4.5 percent) of the total 5,200 crossing took place at unmarked non-
intersection areas. Of these crossings, 60 took place with traffic; 5 of these included the
pedestrian waiting on the median (4 included a traffic flow change to against traffic). The
remaining 172 crossings that took place at an unmarked non-intersection location occurred
against traffic, 77 of which included the pedestrian waiting on the median. (Five included a
traffic flow change to with traffic.)

Pedestrian crossings at the unmarked intersection of Marion Street Northwest and Rhode Island
Avenue Northwest were also recorded. Marion Street is located to the south of Rhode Island
Avenue. Northbound Marion Street traffic must turn (either right or left) when it meets Rhode
Island Avenue. This intersection does not contain a marked pedestrian crossing. There were
2,322 crossings (or 44.7 percent of the 5,200 total crossings) at this unmarked intersection. Of
these, 954 occurred with traffic; 42 included the pedestrian waiting on the median. An additional
84 crossings took place with traffic, but concluded against traffic; 48 of these crossings included
the pedestrian waiting on the median. A further 1,251 crossings at the unmarked intersection
occurred against traffic; 855 of these crossings included the pedestrian waiting on the median.
The remaining 33 crossings involved a traffic flow change where pedestrians began crossing
against traffic and concluded the crossing with traffic; 15 of these crossings included the
pedestrian waiting on the median. In total, 960 (or 41 percent) of the 2,322 crossings that
occurred at the unmarked intersection involved the pedestrian waiting on the median.

Yielding:

In total, there were 105 yielding behaviors (combined across pedestrians and vehicles) at
Location 3. None of these occurred while a pedestrian was in the marked intersection.

There were 11 crossings where pedestrians yielded to vehicles at unmarked non-intersection

locations. There was a single instance where a vehicle yielded to pedestrian at an unmarked non-
intersection area.

In 76 instances, pedestrians yielded to vehicles at the unmarked intersection. In an additional

17 cases, a vehicle yielded to a pedestrian at the unmarked intersection.

47
Evasive Pedestrian Actions:

A total of 113 pedestrians took evasive actions. There were four instances of pedestrian
running/accelerated walking while in the marked intersection during the walk phase. An
additional four pedestrians began crossing during the walk phase but completed their crossing in
the don’t walk phase, who ran/accelerated while in the marked intersection. Nineteen pedestrians
ran/accelerated while crossing at the marked intersection during the don’t walk phase. In
addition, 17 pedestrians ran/accelerated while crossing at unmarked non-intersection locations.
At the unmarked intersection, 61 pedestrians ran/accelerated through the crossing.

In a total of three instances, the pedestrian abruptly stopped to avoid a potential collision with a
vehicle. Of these, two occurred during the don’t walk phase in the marked intersection, and one
took place in the unmarked intersection.

On six occasions, pedestrians made a travel directional change (i.e., returned to the curb). One
occurred in the marked intersection during the walk phase, one occurred at an unmarked non-
intersection location, and the remained four took place at the unmarked intersection.

Evasive Vehicle Actions:

No vehicle evasive actions were recorded during the data collection period at Location 3.

48
 Table 5. Summary of pedestrian crossings at Location 3.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 2,317 — — 4 — 1 — — — —
+traffic flow
20 — — 4 — — — — — —
change
Don’t Walk 298 — — 19 2 — — — — —
+traffic flow
11 — — — — — — — — —
change
Sum 2,646 — — 27 2 1 — — — —
Unmarked Non-Intersection
With Traffic 52 — — 2 — — — — — —
+traffic flow
8 — — — — — — — — —
change
Against Traffic 164 9 1 14 — 1 — — — —
+traffic flow
8 2 — 1 — — — — — —
change
Sum 232 11 1 17 — 1 — — — —
Unmarked Intersection
With Traffic 954 3 1 9 — — — — — —
+traffic flow
84 3 3 5 — 1 — — — —
change
Against Traffic 1,251 67 13 47 1 3 — — — —
+traffic flow
33 3 — — — — — — — —
change
Sum 2,322 76 17 61 1 4 — — — —
Location 3 Total 5,200 87 18 105 3 6 — — — —
— Indicates no actions of this type were recorded.

49
Location 4

Total Crossings:

There were a total of 13,199 pedestrian crossings on I (eye) Street Northwest between 13th and
14th Streets northwest during the coded data collection times. (See table 6 for a summary of
crossings.) Of these crossings, 12,034 (or 91.2 percent) occurred within the marked intersection
(i.e., marked crosswalk). These crossings can be broken down into when they occurred during
the light phases. During the walk phase, 8,973 pedestrians crossed within the near marked
intersection. An additional 2,361 crossings occurred in the marked intersection during the don’t
walk phase. In addition, 700 crossings occurred in the marked intersection and included a traffic
flow change. Of these crossings, 94 began in the walk phase and ended in the don’t walk phase.
An additional 606 began in the don’t walk phase and concluded in the walk phase. In other
words, pedestrians were more likely to begin their crossing during the don’t walk phase than the
walk phase.

At Location 4, 1,165 (or 8.8 percent) of the total 13,199 crossings took place at unmarked non-
intersection areas. Of these crossings, 710 took place with traffic, and 365 took place against
traffic. A portion of the pedestrian crossings that took place at an unmarked non-intersection area
involved a traffic flow change. Twenty-two of these crossings began with traffic and ended
against traffic, and 68 began against traffic and ended with traffic.

Yielding:

In total, there were 2,693 yielding behaviors (combined across pedestrians and vehicles) at
Location 4. In 15 cases, the pedestrian yielded to the vehicle in the marked intersection during
the walk phase. In 2,681 instances, a vehicle yielded to a pedestrian in the marked intersection
during the walk phase.

In addition, there were 11 instances of yielding behavior while pedestrians crossed during a
traffic flow change. In six of these cases, the pedestrians began crossing during the don’t walk
phase and completed the crossing during the walk phase and vehicles yielded to the pedestrian.
In three cases, pedestrians yielded to vehicles after beginning to cross during the don’t walk
phase and completing the crossing the walk phase. In the remaining two instances, pedestrians
that began their crossing during the walk phase and completed it during the don’t walk phase
yielded to vehicles in the marked intersection.

Evasive Pedestrian Actions:

A total of 288 pedestrians took evasive actions. There were 136 instances of pedestrian
running/accelerated walking while in the marked intersection during the walk phase. An
additional 12 pedestrians who began crossing during the walk phase but completed their crossing
in the don’t walk phase, ran/accelerated while in the marked intersection. During the don’t walk
phase, 104 pedestrians ran/accelerated while crossing at the marked intersection. An additional
six pedestrians began crossing during the don’t walk phase but completed during the walk phase
while running/accelerated walking. While in the unmarked non-intersection, 16 pedestrians
ran/accelerated while completing their crossing.

50
In a total of 10 instances, the pedestrian abruptly stopped to avoid a potential collision with a
vehicle. Of these, four occurred during the walk phase in the marked intersection. Three
crossings occurred entirely during the don’t walk phase. An additional three pedestrians abruptly
stopped in the marked intersection after beginning crossing during the don’t walk phase and
completed during the walk phase.

In five instances, pedestrians made a travel directional change. Of these, three occurred in the
marked intersection during the walk phase, one occurred in the marked intersection during the
don’t walk phase, and one occurred in an unmarked non-intersection location.

Evasive Vehicle Actions:

No vehicle evasive actions were recorded during the data collection period at Location 4.

51
 Table 6. Summary of pedestrian crossings at Location 4.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 8,973 15 2,681 136 4 3 — — — —
+traffic flow
94 2 12 12 — — — — — —
change
Don’t Walk 2,361 — — 104 3 1 — — — —
+traffic flow
606 3 6 6 3 — — — — —
change
Sum 12,034 20 2,687 258 10 4 — — — —
Unmarked Non-Intersection
With Traffic 710 1 5 9 — 1 — — — —
+traffic flow
22 — — — — — — — — —
change
Against Traffic 365 — 1 7 — — — — — —
+traffic flow
68 — — — — — — — — —
change
Sum 1,165 1 6 16 — 1 — — — —
Location 4 Total 13,199 21 2,693 274 10 5 — — — —
— Indicates no actions of this type were recorded.

52
Location 5

Total Crossings:

There were a total of 10,635 pedestrian crossings on H Street Northwest between 14th and
15th Streets Northwest during the coded data collection times. (See table 7 for a summary of
crossings.) Location 5 was not coded after February 13 because the camera was facing a different
direction. Of the coded crossings, 10,032 (or 94.3 percent) occurred within the marked
intersection (i.e., marked crosswalk). These crossings can be broken down into when they
occurred during the light phases. During the walk phase, 9,936 pedestrians crossed within the
near marked intersection. An additional 52 crossings occurred in the marked intersection during
the don’t walk phase. In addition, 44 crossings occurred in the marked intersection that included
a traffic flow change. Of these, 39 began in the walk phase and ended in the don’t walk phase.
Five crossings began in the don’t walk phase and concluded in the walk phase.

At Location 5, 603 (or 5.7 percent) of the total 10,635 pedestrian crossing took place at
unmarked non-intersection areas. Of these crossings, 552 took place with traffic, and 35 took
place against traffic. A portion of the pedestrian crossings that took place at an unmarked non-
intersection area involved a traffic flow change. Fifteen of these crossings began with traffic and
ended against traffic, while one began against traffic and ended with traffic.

Yielding:

There were a total of 19 yielding behaviors (combined across pedestrians and vehicles) at
Location 5. In three cases, pedestrians yielded to vehicles, and in a single case, a vehicle yielded
to a pedestrian in the marked intersection during the walk phase. In addition, in 11 instances a
vehicle yielded to a pedestrian in the marked intersection that involved a traffic flow change; in
each, the pedestrian began his or her crossing during the walk phase and concluded during the
don’t walk phase.

In two instances, a pedestrian yielded to a vehicle, and in two instances, a vehicle yielded to a
pedestrian in an unmarked non-intersection area.

Evasive Pedestrian Actions:

A total of 235 pedestrians took evasive actions. There were 163 instances of pedestrian running/
accelerated walking while in the marked intersection during the walk phase. An additional
29 pedestrians, who began crossing during the walk phase but completed their crossing in the
don’t walk phase, ran/accelerated while in the marked intersection. During the don’t walk phase,
13 pedestrians ran/accelerated while crossing at the marked intersection. An additional two
pedestrians, who began crossing during the don’t walk phase but completed their crossing in the
walk phase, ran/accelerated while in the marked intersection. Nineteen pedestrians
ran/accelerated while crossing in an unmarked non-intersection area.

In a total of five instances, the pedestrian abruptly stopped to avoid a potential collision with a
vehicle. Of these, two occurred during the walk phase in the marked intersection. Two crossings
occurred during the don’t walk phase, and a single pedestrian abruptly stopped while crossing at
an unmarked non-intersection location.

53
In eight instances, pedestrians made a travel directional change. Of these, four occurred in the
marked intersection during the walk phase, and four occurred in the marked intersection during
the don’t walk phase.

Evasive Vehicle Actions:

No vehicle evasive actions were recorded during the data collection period at Location 5.

54
 Table 7. Summary of pedestrian crossings at Location 5.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 9,936 3 1 163 2 4 — — — —
+traffic flow
39 — 11 29 — — — — — —
change
Don’t Walk 52 — — 13 2 4 — — — —
+traffic flow
5 — — 2 — — — — — —
change
Sum 10,032 3 12 207 4 8 — — — —
Unmarked Non-Intersection
With Traffic 552 1 1 11 1 — — — — —
+traffic flow
15 — 1 3 — — — — — —
change
Against Traffic 35 1 — 5 — — — — — —
+traffic flow
1 — — — — — — — — —
change
Sum 603 2 2 19 1 — — — — —
Location 5 Total 10,635 5 14 226 5 8 — — — —
— Indicates no actions of this type were recorded.

55
Location 6

Total Crossings:

There were a total of 16,418 pedestrian crossings on 14th Street Northwest between New York
Avenue Northwest and G Street Northwest during the coded data collection times. (See table 8
for a summary of crossings.) Of these crossings, 15,312 (or 93.3 percent) occurred within the
marked intersection. These crossings can be broken down into when they occurred during the
light phases. During the walk phase, 14,879 pedestrians crossed within the marked intersection.
An additional 151 crossings occurred in the marked intersection during the don’t walk phase. In
addition, 282 crossings occurred in the marked intersection and included a traffic flow change.
Of these, 135 began in the walk phase and ended in the don’t walk phase. An additional
147 crossings began in the don’t walk phase and concluded in the walk phase.

At Location 6, 1,106 (or 6.7 percent) of the total 16,418 crossings took place at unmarked non-
intersection areas. Of these crossings, 767 took place with traffic, and 183 took place against
traffic. A portion of the pedestrian crossings that took place at an unmarked non-intersection area
involved a traffic flow change. Forty-five of these crossings began with traffic and ended against
traffic, while 111 began against traffic and ended with traffic.

Yielding:

There were a total of 1,357 yielding behaviors (combined across pedestrians and vehicles) at
Location 6. In 22 cases, pedestrians yielded to vehicles, and in 1,300 cases, vehicles yielded to
pedestrians in the marked intersection during the walk phase. There were also eight instances of
yielding that involved a traffic flow change in the marked intersection. In a single instance, a
vehicle yielded to the pedestrian, who began the crossing during the walk phase and concluded
during the don’t walk phase in the marked intersection. In seven cases, the pedestrian yielded to
a vehicle in a crossing that began during the don’t walk phase of and concluded during the walk
phase.

In 26 cases, pedestrians yielded to vehicles in an unmarked non-intersection area. In a single

instance, a vehicle yielded to a pedestrian in an unmarked non-intersection.

Evasive Pedestrian Actions:

A total of 416 pedestrians took evasive actions. There were 263 instances of pedestrian running/
walking while in the marked intersection during the walk phase. An additional 38 pedestrians,
who began crossing during the walk phase but completed their crossing in the don’t walk phase,
ran/accelerated while in the marked intersection. During the don’t walk phase, 31 pedestrians
ran/accelerated while crossing at the marked intersection. An additional eight pedestrians, who
began crossing during the don’t walk phase but completed their crossing in the walk phase,
ran/accelerated while in the marked intersection. Fifty-eight pedestrians ran/accelerated while
crossing in an unmarked non-intersection area.

In a total of 13 instances, the pedestrian abruptly stopped to avoid a potential collision with a
vehicle. Of these, four occurred during the walk phase, and four occurred during the don’t walk
phase in the marked intersection. Two additional cases involved a traffic flow change;

56
pedestrians began their crossing during the don’t walk phase and concluded during the walk
phase. Three pedestrians abruptly stopped while crossing at an unmarked non-intersection
location.

In 11 instances, pedestrians made a travel directional change. Of these, six occurred in the
marked intersection during the walk phase, and four occurred in the marked intersection during
the don’t walk phase. A single directional change instance involved a traffic flow change; the
pedestrian began crossing during the don’t walk phase and concluded during the walk phase.

Evasive Vehicle Actions:

No vehicle evasive actions were recorded during the data collection period at Location 6.

57
 Table 8. Summary of pedestrian crossings at Location 6.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 14,879 22 1,300 263 4 6 — — — —
+traffic flow
135 — 1 38 — — — — — —
change
Don’t Walk 151 — — 31 4 4 — — — —
+traffic flow
147 7 — 8 2 1 — — — —
change
Sum 15,312 29 1,301 340 10 11 — — — —
Unmarked Non-Intersection
With Traffic 767 5 1 22 1 — — — — —
+traffic flow
45 1 — 11 — — — — — —
change
Against Traffic 183 15 — 21 — — — — — —
+traffic flow
111 5 — 4 2 — — — — —
change
Sum 1,106 26 1 58 3 — — — — —
Location 6 Total 16,418 55 1,302 398 13 11 — — — —
— Indicates no actions of this type were recorded.

58
Location 7

Total Crossings:

There were a total of 12,958 crossings on 19th Street Northwest between Pennsylvania Avenue
Northwest and H Street Northwest during the coded data collection times. (See table 9 for a
summary of crossings.) Of these crossings, 12,567 (or 97.0 percent) occurred within the marked
intersection. These crossings can be broken down into when they occurred during the light
phases. During the walk phase, 9,778 pedestrians crossed within the marked intersection. An
additional 2,422 crossings occurred in the marked intersection during the don’t walk phase. In
addition, 367 crossings occurred in the marked intersection and included a traffic flow change.
Of these, 169 began in the walk phase and ended in the don’t walk phase. A further
198 crossings began in the don’t walk phase and concluded in the walk phase.

At Location 7, 391 (or 3.0 percent) of the total 12,958 crossings took place at unmarked non-
intersection areas. Of these crossings, 235 took place with traffic, and 155 took place against
traffic. Four crossings included a traffic flow change. A single pedestrian crossing began with
traffic and ended against traffic, while three crossings began against traffic and ended with
traffic.

Yielding:

There were a total of 5,100 yielding behaviors (combined across pedestrians and vehicles) at
Location 7. In 72 instances, pedestrians yielded to vehicles, and in 4,958 cases, a vehicle yielded
to a pedestrian in the marked intersection during the walk phase. There were also four cases
where a pedestrian yielded to a vehicle and 15 cases where a vehicle yielded to a pedestrian in
the marked intersection during the don’t walk phase.

There were 37 instances of yielding that involved a traffic flow change in the marked
intersection. In 28 instances, a vehicle yielded to the pedestrian, who began the crossing during
the walk phase and concluded during the don’t walk phase in the marked intersection. In seven
cases, the pedestrian yielded to a vehicle in a crossing that began during the walk phase and
concluded during the don’t walk phase. There were two instances of yielding for pedestrians who
began crossing during the don’t walk phase and concluded during the walk phase; once the
pedestrian yielded to the vehicle and once the vehicle yielded to the pedestrian.

In two cases, pedestrians yielded to vehicles in an unmarked non-intersection area. In

12 instances, a vehicle yielded to a pedestrian in an unmarked non-intersection.

Evasive Pedestrian Actions:

A total of 264 pedestrians took evasive actions. There were 22 instances of pedestrian
running/accelerated walking while in the marked intersection during the walk phase. An
additional 25 pedestrians, who began crossing during the walk phase but completed their
crossing in the don’t walk phase, ran/accelerated while in the marked intersection. During the
don’t walk phase, 126 pedestrians ran/accelerated while crossing at the marked intersection.
Thirty-two pedestrians ran/accelerated while crossing in an unmarked non-intersection area.

59
In a total of 51 instances, the pedestrian abruptly stopped to avoid a potential collision with a
vehicle. Of these, 49 occurred during the don’t walk phase in the marked intersection. In a single
case that involved a traffic flow change, the pedestrian began crossing during the don’t walk
phase and concluded during the walk phase. A single pedestrian abruptly stopped while crossing
at an unmarked non-intersection location.

In 31 instances, pedestrians made a travel directional change. Of these, 29 occurred in the

marked intersection during the don’t walk phase. A single directional change instance involved a
pedestrian who began crossing during the walk phase in the marked intersection. A single
pedestrian had a directional change while crossing at an unmarked non-intersection location.

Evasive Vehicle Actions:

No vehicle evasive actions were recorded during the data collection period at Location 7.

60
 Table 9. Summary of pedestrian crossings at Location 7.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 9,778 72 4,958 22 — — — — — —
+traffic flow
169 7 28 25 — — — — — —
change
Don’t Walk 2,422 4 15 126 49 — — — — —
+traffic flow
198 1 1 — 1 — — — — —
change
Sum 12,567 84 5,002 173 50 — — — — —
Unmarked Non-Intersection
With Traffic 235 1 5 10 — — — — — —
+traffic flow
1 — — — — — — — — —
change
Against Traffic 155 1 7 22 1 — — — — —
+traffic flow
3 — — — — — — — — —
change
Sum 391 2 12 32 1 — — — — —
Location 7 Total 12,958 86 5,014 205 51 — — — — —
— Indicates no actions of this type were recorded.

61
Location 8

Total Crossings:

There were a total of 1,574 crossings on Connecticut Avenue between Oliver Street and
Northampton Street Northwest during the coded data collection times. (See table 10 for a
summary of crossings.) Of these crossings, 1,209 (or 76.8 percent) occurred within the marked
intersection. These crossings can be broken down into when they occurred during the light
phases. During the walk phase, 1,168 pedestrians crossed within the marked intersection. An
additional 21 crossings occurred in the marked intersection during the don’t walk phase. In
addition, 20 crossings occurred in the marked intersection and included a traffic flow change. Of
these, 10 began in the walk phase and ended in the don’t walk phase. An additional 10 crossings
began in the don’t walk phase and concluded in the walk phase.

At Location 7, 365 (or 23.2 percent) of the total 1,574 crossings took place at unmarked non-
intersection areas. Of these crossings, 261 took place with traffic, and 86 took place against
traffic. A portion of the pedestrian crossings that took place at an unmarked non-intersection area
involved a traffic flow change. Ten of these crossings began with traffic and ended against
traffic, while eight began against traffic and ended with traffic.

In 12 instances, a pedestrian waited in the median to complete a crossing at Location 8. Three of

these occurred when pedestrians crossed during the don’t walk phase in the marked intersection.
There were also nine crossings involving waiting on the median that occurred in the marked
intersection and included a traffic flow change. Two of these crossings began with traffic and
ended against traffic, and six began against traffic and ended with traffic. In a single case, a
pedestrian waited in the median area when crossing at an unmarked non-intersection location.

Yielding:

There were a total of 12 yielding behaviors (combined across pedestrians and vehicles) at
Location 8. In a single instance, a pedestrian yielded to a vehicle, and in four cases, a vehicle
yielded to a pedestrian in the marked intersection during the walk phase. Two instances of
yielding involved a traffic flow change in the marked intersection. In both instances, a vehicle
yielded to the pedestrian who began the crossing during the walk phase and concluded during the
don’t walk phase in the marked intersection.

In a single case, a pedestrian yielded to a vehicle in an unmarked non-intersection area. In four

instances, a vehicle yielded to a pedestrian in an unmarked non-intersection.

Evasive Pedestrian Actions:

A total of 136 pedestrians took evasive actions at Location 8. There were 40 instances of
pedestrian running/accelerated walking while in the marked intersection during the walk phase.
An additional seven pedestrians, who began crossing during the walk phase but completed their
crossing in the don’t walk phase, ran/accelerated while in the marked intersection. During the
don’t walk phase, eight pedestrians ran/accelerated while crossing at the marked intersection. An
additional four pedestrians began crossing during the don’t walk phase, but completed their

62
crossing in the walk phase, ran/accelerated while in the marked intersection. Seventy-two
pedestrians ran/accelerated while crossing in an unmarked non-intersection area.

In a total of eight instances, the pedestrian abruptly stopped to avoid a potential collision with a
vehicle. Of these, one occurred during the don’t walk phase in the marked intersection. In three
cases that involved a traffic flow change, the pedestrian began crossing during the don’t walk
phase and concluded during the walk phase. Four pedestrians abruptly stopped while crossing at
an unmarked non-intersection location.

There was a single instance of a pedestrian directional change. In this case, there was a traffic
flow change where the pedestrian began crossing during the walk phase and completed the
crossing during the don’t walk phase.

Evasive Vehicle Actions:

No vehicle evasive actions were recorded during the data collection period at Location 8.

63
 Table 10. Summary of pedestrian crossings at Location 8.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 1,168 1 4 40 — — — — — —
+traffic flow
10 — 2 7 — 1 — — — —
change
Don’t Walk 21 — — 8 1 — — — — —
+traffic flow
10 — — 4 3 — — — — —
change
Sum 1,209 1 6 59 4 1 — — — —
Unmarked Non-Intersection
With Traffic 261 — 1 36 — — — — — —
+traffic flow
10 1 1 4 — — — — — —
change
Against Traffic 86 — 2 31 4 — — — — —
+traffic flow
8 — — 1 — — — — — —
change
Sum 365 1 4 72 4 — — — — —
Location Total 1,574 2 10 131 8 1 — — — —
— Indicates no actions of this type were recorded.

64
Predictive Model

It was hoped that the results of this study might be used to develop a model that would predict
whether a pedestrian would cross at a marked intersection or an unmarked non-intersection based
on the features of the environment. Several approaches were taken to model the data.

Location 3 contained a significant proportion of crossings at the unmarked intersection, which

made it significantly different than the other seven data collection areas (which did not include
this type of crossing). As a result, Location 3 was excluded from all modeling presented in this
section.

Training and testing data sets were created from the raw data. The training data is used to build
predictions about the testing data. (All data are drawn from the original raw data set of
65,725 crossings.) To generate the training data set, a stratified random sample without
replacement was performed using PROC SURVEYSELECT in SAS®. Location (i.e., marked
intersection or unmarked non-intersection) was the stratifying variable. Approximately
70 percent of the raw data was selected for the training data set, and the remaining 30 percent
was assigned to the testing data set.

Of the 42,231 observations selected for the training data set, about 93 percent involved
pedestrians crossing at marked intersections. Hence, unmarked non-intersection crossings were
considered a rare event. One technique to handle the occurrence of rare events is to over sample
the rare events in the training data set. This methodology was used here.

Ten subsets of the training data set were created. Each subset was designed so that 50 percent of
the observations involved marked intersection crossings and the remaining 50 percent of the
observations involved unmarked non-intersection crossings. All 2,893 unmarked non-
intersection crossings from the original training set were included in each training subset. To
generate the remaining observations for each subset, 2,893 observations were randomly selected
using a simple random sample without replacement from the 39,338 marked intersection
crossings in the original training set. This process did not involve stratifying by location. Each
subset contained a different set of marked intersection crossings. Thus, each training subset
contained 5,786 observations.

PROC GLMSELECT was used to model the pedestrian crossing location. Although the crossing
location was a binary variable (i.e., only two possible outcomes), previous research has shown
that appropriate linear model selection techniques can produce models whose prediction
capabilities are competitive to those produced through logistic modeling.(27) Each training subset
was used once to model the pedestrian crossing location so that 10 models were produced. The
testing data set was used to test each model. Under each model, 33.33 percent of the training
subset was reserved for model validation. Stepwise selection was performed. The adjusted
r-square was used for selecting and stopping criteria. The average squared error of the validation
data was used as the choosing criterion for the model to prevent over-fitting of the training data.
Hierarchy was assumed for all model effects, meaning that an interaction term could not enter
the model unless all main effects were already present in the model. Similarly, main effects
could not exit the model before any respective interaction effects. Available predictors for each
of the 10 models were A through O (see table 11; further details of each location are provided

65
under each respective location description) and all second-order interactions, for a total of 121
effects. Predicted probabilities less than or equal to 0.5 corresponded to a predicted binary value
of 0 (i.e., unmarked non-intersection). All other probabilities corresponded to a predicted binary
value of 1 (i.e., marked intersection).

Table 11. Predictors and their respective descriptions used for the models.
Label Description Coding
A Distance to the next marked crosswalk Distance in ft
B AADT Expressed in thousands and rounded to the
nearest 100
C One-way or two-way street 1 or 2
D Presence of physical barriers that might No barrier (0), partial barrier (1), or mostly
prevent a pedestrian from crossing the blocked/large barrier (2)
roadway
E Presence of a bus stop None (0), bus exit near marked intersection
(1), bus exit at non-intersection (2)
F Range of the number of trip originators/ Range from very few (1) to a lot (5)
destinations
G Presence of parking along the roadway Yes (1) or no (0)
H Presence of a center turning lane Yes (1) or no (0)
I Presence of a right turn only turning lane Yes (1) or no (0)
J Length of walk phase Time in s
K Length of don’t walk phase Time in s
L Curb-to-curb distance Distance in ft
M Presence and type of median No median (0), soft (1), hard (2), median
only on one side of crosswalk (3)
N Presence of cross streets between marked No cross street, light-controlled cross
crosswalks street, not light-controlled cross street
O Far marked crosswalk light controlled Yes (1) or no (0)
Note: Values in parentheses are the values assigned to categorical variables.

The following subsections present the 10 (statistically) chosen models. For categorical variables,
the second value corresponds to the value of the variable. For instance, xC,1 represents value 1 of
predictor C (i.e., one-way traffic direction).

66
Model 1

 Chosen Effects: Intercept, XC, XD, XI, and XM.

 Adjusted r-square: 0.0551.
 Model:
CrossingLoc = 0.1932 – 0.504xC,1 + 0.1385xD,0 + 0.4082xD,1 – 0.1265xI,2 + 0.2863xM,0 +
0.1102xM,2.
 Classification of Training Subset: 58.30 percent correct.
 Classification of Testing Data Set: 42.60 percent correct.
o Of those correctly assigned, 87.91 percent were for marked intersection crossings.
o Of those incorrectly assigned, 97.12 percent were for marked intersection crossings.

Model 2

 Chosen Effects: Intercept, XA, XD, XI, and XM.

 Adjusted r-square: 0.0603.
 Model:
CrossingLoc = 0.1491 – 0.0000xA + 0.1301xD,0 + 0.3542xD,1 – 0.0933xI,0 + 0.3138xM,0 +
0.1956xM,2.
 Classification of Training Subset: 58.62 percent correct.
 Classification of Testing Data Set: 42.60 percent correct.
o Of those correctly assigned, 87.91 percent were for marked intersection crossings.
o Of those incorrectly assigned, 97.12 percent were for marked intersection crossings.

Model 3

 Chosen Effects: Intercept, XD, XI, XJ, and XM.

 Adjusted r-square: 0.0614.
 Model:
CrossingLoc = 0.8614 + 0.3906xD,0 + 0.9442xD,1 – 0.6139xI,0 – 0.0234xJ + 0.5613xM,0 +
0.2649xM,2.
 Classification of Training Subset: 58.95 percent correct.
 Classification of Testing Data Set: 42.60 percent correct.
o Of those correctly assigned, 87.91 percent were for marked intersection crossings.
o Of those incorrectly assigned, 97.12 percent were for marked intersection crossings.

Model 4

 Chosen Effects: Intercept, XD, XE, and XM.

 Adjusted r-square: 0.0524.
 Model:
CrossingLoc = 0.0217 + 0.0806xD,0 + 0.3239xD,1 + 0.0642xE,0 + 0.3555xM,0 + 0.2333xM,2.
 Classification of Training Subset: 58.07 percent correct.
 Classification of Testing Data Set: 66.28 percent correct.
o Of those correctly assigned, 94.89 percent were for marked intersection crossings.
o Of those incorrectly assigned, 89.88 percent were for marked intersection crossings.

67
Model 5

 Chosen Effects: Intercept, XA, XD, XI, and XM.

 Adjusted r-square: 0.0560.
 Model:
CrossingLoc = 0.1856 – 0.0001xA + 0.1312xD,0 + 0.3366xD,1 – 0.0812xI,0 + 0.2868xM,0 +
0.1745xM,2.
 Classification of Training Subset: 58.14 percent correct.
 Classification of Test Data Set: 42.60 percent correct.
o Of those correctly assigned, 87.91 percent were for marked intersection crossings.
o Of those incorrectly assigned, 97.12 percent were for marked intersection crossings.

Model 6

 Chosen Effects: Intercept, XD, XI, XJ, and XM.

 Adjusted r-square: 0.0588.
 Model:
CrossingLoc = 1.1831 + 0.5053xD,0 + 1.1918xD,1 – 0.8352xI,0 – 0.0354xJ + 0.7382xM,0 +
0.3473xM,2.
 Classification of Training Subset: 57.93 percent correct.
 Classification of Test Data Set: 42.60 percent.
o Of those correctly assigned, 87.91 percent were for marked intersection crossings.
o Of those incorrectly assigned, 97.12 percent were for marked intersection crossings.

Model 7

 Chosen Effects: Intercept, XC, XD, XI, and XM.

 Adjusted r-square: 0.0584.
 Model:
CrossingLoc = 0.1559 – 0.0453xC,1 + 0.1663xD,0 + 0.4348xD,1 – 0.1771xI,0 + 0.2877xM,0 +
0.0888xM,2.
 Classification of Training Subset: 58.38 percent correct.
 Classification of Test Data Set: 42.60 percent correct.
o Of those correctly assigned, 87.91 percent were for marked intersection crossings.
o Of those incorrectly assigned, 97.12 percent were for marked intersection crossings.

Model 8

 Chosen Effects: Intercept, XD, XI, XJ, and XM.

 Adjusted r-square: 0.0476.
 Model:
CrossingLoc = 1.4533 + 0.5920xD,0 + 1,3440xD,1 – 1.0218xI,0 – 0.0426xJ + 0.7929xM,0 +
0.3502xM,2.
 Classification of Training Subset: 57.86 percent correct.

68
  Classification of Testing Data Set: 42.60 percent correct.
o Of those correctly assigned, 87.91 percent were for marked intersection crossings.
o Of those incorrectly assigned, 97.12 percent were for marked intersection crossings.

Model 9

 Chosen Effects: Intercept, XD, XI, XJ, and XM.

 Adjusted r-square: 0.0519.
 Model:
CrossingLoc = 1.6741 + 0.7527xD,0 + 1.6295xD,1 – 1.2252xI,0 – 0.0524xJ +
0.9266xM,0 + 0.4021xM,2.
 Classification of Training Subset: 58.37 percent correct.
 Classification of Testing Data Set: 66.28 percent correct.
o Of those correctly assigned, 94.89 percent were for marked intersection crossings.
o Of those incorrectly assigned, 89.88 percent were for marked intersection crossings.

Model 10

 Chosen Effects: Intercept, XC, XD, XI, and XM.

 Adjusted r-square: 0.0546.
 Model:
CrossingLoc = 0.1803 – 0.0506xC,1 + 0.1533xD,0 + 0.4221xD,1 – 0.1205xI,0 + 0.2801xM,0 +
0.1033xM,2.
 Classification of Training Subset: 58.16 percent correct.
 Classification of Testing Data Set: 42.60 percent correct.
o Of those correctly assigned, 87.91 percent were for marked intersection crossings.
o Of those incorrectly assigned, 97.12 percent were for marked intersection crossings.

Model Summary

Of the 121 effects entered into each of the 10 models, only XA, XC, XD, XE, XI, XJ, and XM
were ever chosen. No interactions were selected for the models using the aforementioned model
restrictions. Table 12 summarizes each of the chosen effects and their respective parameter
estimates.

69
 Table 12. Summary of each of the model-selected effects (predictors) and their parameter
estimates.
Occurrences
Effect Mean Minimum Maximum in Model
Intercept 0.6058 0.0217 1.6741 10
XA -0.0001 -0.0001 0.0000 2
XC,1 -0.0488 -0.0506 -0.0453 3
XD,0 0.3041 0.0806 0.7527 10
XD,1 0.7389 0.3239 1.6295 10
XE 0.0642 0.0642 0.0642 1
XI,0 -0.4705 -1.2252 -0.0812 9
XJ -0.0385 -0.0524 -0.0234 4
XM,0 0.4829 0.2801 0.9266 10
XM,2 0.2290 0.0888 0.4021 10

XD (the presence of a physical barrier) and XM (the presence of a median) were selected for
every model. Each of these variables had three values, two of which were assigned parameter
estimates. For XD, value 0 (no barrier) always yielded a smaller parameter estimate than value 1
(partial barrier/blocked crosswalk). For XM, value 2 (median only on one side of the crosswalk)
always yielded a smaller parameter estimate than value 0 (no median). XI,0 (no dedicated right
turn only lane) was chosen in 9 of the 10 models. Parameter estimates for XC (traffic
directionality), XI (dedicated right turn only lane), and XJ (length of walk phase) were always
negative, indicating that these effects tended to decrease the probability of a marked intersection
crossing (i.e., increase the probability of an unmarked non-intersection crossing). The mean
intercept parameter estimate was 0.6058 (greater than 0.5), indicating that baseline predictions
yielded a marked intersection crossing. Although XA (distance to the next marked crossing) was
selected for the model twice, the influence of this variable was negligible in both cases.

Each of the 10 models accurately predicted the testing data set only about 50 percent of the time.
Furthermore, the largest adjusted r-square value was only .0614. In other words, the model
explained only about 6.14 percent of the variance in the data. These two factors combined
suggest that any single model calculated thus far is not sufficient to predict whether pedestrians
are more likely to cross at a marked intersection than at an unmarked non-intersection. As a
result, more detailed and site- and factor-specific human factors analyses are described in the
next section. It is hoped that this will provide greater insight regarding which environmental
factors influence pedestrian crossing behavior.

Factor Specific Analyses

This results section evaluates each of the recorded crossing variables in detail. Although there
are too few data collection sites in phase 1 to use inferential statistics to compare one site with
another, general trends are examined.

70
Crossing Location

Table 13 summarizes the percentage of pedestrians, by location, who crossed at marked

intersections, unmarked non-intersections, and unmarked intersections. The percentage of each
type of crossing is presented at each location. For example, at Location 1, 75.13 percent of the
total crossings took place in the marked intersection during the walk phase. When all locations
are combined, the mean percentage of pedestrians who crossed at the marked intersection is
85.35. However, the Location 3 marked intersection percentage of 50.9 is 2.27 standard
deviations below the mean, making it an outlier.

Further examination of Location 3 shows that the relatively low number of pedestrians crossing
at the marked intersection is the result of pedestrians crossing at the unmarked intersection. In
fact, pedestrians were about equally likely to cross in the marked intersection (50.88 percent) as
the unmarked intersection (44.65 percent).

If Location 3 is removed, the mean percentage of pedestrians who crossed at the marked
intersection in the remaining seven locations is 90.27. Here, the Location 8 marked intersection
value of 76.8 percent is 2.05 standard deviations below the mean, making it an outlier. The
Location 7 marked intersection crossing value of 97.0 percent is 1.02 standard deviations above
the mean. While this is not an outlier, it is the most extreme high value. As a result, this is
explored further in the subsequent discussion section.

Because of the third crossing area choice at Location 3 (i.e., the unmarked intersection), it is
worthwhile to further explore the percentage of pedestrians who crossed at unmarked non-
intersection areas. Not surprisingly, the Location 8 unmarked non-intersection value of
28.2 percent is 2.32 standard deviations above the mean and consequently is classified as an
outlier. If the Location 8 value is removed, the mean percentage of pedestrians crossing at an
unmarked non-intersection at the remaining seven locations is reduced to 7.06. No further
outliers exist. However, Location 1 and Location 7 remain the two most extreme values at
1.63 standard deviations above the mean and 1.34 standard deviations below the mean,
respectively.

Further exploration for the possible reasons for these differences is presented in more detail in
the phase 1 discussion section.

71
 Table 13. Percentage of pedestrians at each crossing area in each data collection location.
Location
1 2 3 4 5 6 7 8
Marked Intersection
Walk 75.13 87.91 44.56 67.98 93.43 90.63 75.46 74.21
+traffic flow change .72 1.22 .38 .71 .37 .82 1.30 .64
Don’t Walk 8.47 1.75 5.73 17.89 .49 .92 18.69 1.33
+traffic flow change 3.60 .52 .21 4.59 .05 .90 1.53 .64
Overall in Intersection 87.93 91.30 50.88 91.17 94.33 93.26 96.98 76.81
Unmarked Non-Intersection
With Traffic 5.23 6.24 1.00 5.38 5.19 4.67 1.81 16.58
+traffic flow change .45 .30 .15 .17 .14 .27 .01 .64
Against Traffic 5.77 1.60 3.15 2.77 .33 1.11 1.20 5.46
+traffic flow change .63 .56 .15 .52 .01 .68 — .51
Overall in Unmarked Non-
12.07 8.70 4.46 8.83 5.67 6.74 3.02 23.19
Intersection
Unmarked Intersection
With Traffic — — 18.34 — — — — —
+traffic flow change — — 1.62 — — — — —
Against Traffic — — 24.06 — — — — —
+traffic flow change — — .63 — — — — —
Overall in Unmarked
— — 44.65 — — — — —
Intersection
— Indicates no actions of this type were recorded.

Pedestrian Yielding

Table 14 summarizes the percentage of pedestrians who yielded to vehicles within each crossing
type. As an example, at Location 1, of the 834 pedestrians who crossed during the walk phase in
the marked intersection, 1 pedestrian yielded to a vehicle. This is equivalent to .12 percent. Also
at Location 1, of the total 976 crossings in the marked intersection, only 1 pedestrian yielded to a
vehicle. This is the equivalent of .10 percent of the total crossings in the marked intersection.

Overall, the mean percentage of pedestrians who yielded to vehicles was .53. If all of the
locations are compared with one another, the Location 3 value of 1.67 percent was 2.15 standard
deviations above the mean and is considered an outlier. If Location 3 is removed from these
values, no other outliers remain. These overall yielding behaviors do not provide much
information about where pedestrians are yielding to vehicles. As a result, yielding behaviors by
crossing area are examined next.

The mean percentage of pedestrians who yielded to vehicles in the marked intersection was .19.
If all of the locations are compared with one another, the Location 7 value of .67 percent was
2.42 standard deviations above the mean and is considered an outlier. If Location 7 is removed,
no further outliers exist.

72
Next, the mean percentage of pedestrians who yielded to vehicles in unmarked non-intersections
was examined. Overall, the mean percentage of yielding was 2.19. No outliers exist here. In
other words, when looking at these eight locations alone, none are significantly different with
regard to the proportion of pedestrians who yielded to vehicles in unmarked non-intersection
areas.

A t-test was performed to determine whether the proportion of pedestrian yielding behaviors
differed between the marked intersection and the unmarked non-intersection.

A significant difference between the two locations was found, t(7) = -2.40, p = .048. Crossings
where the pedestrian crossed entirely during the walk phase in the marked intersection were
compared with all other crossings made where the pedestrian yielded. There was no significant
difference in the percentage of pedestrians who yielded to vehicles during the walk phase in the
marked intersection (M = .18 percent) compared with those pedestrians who yielded to vehicles
in all other situations (M = 1.68 percent), t(7) = -2.11, p > .05.

Table 14. Percentage of pedestrians yielding to vehicles in each crossing area at each data
collection location.
Location
1 2 3 4 5 6 7 8
Marked Intersection
Walk .12 .12 — .18 .03 .15 .74 .09
+traffic flow change — 1.92 — 2.12 — — 4.14 —
Don’t Walk — — — 0 — — .17 —
+traffic flow change — 25.00 — .50 — 4.76 .51 —
Overall Yielding in
.10 .28 — .17 .03 .19 .67 .08
Intersection
Unmarked Non-Intersection
With Traffic 3.45 2.08 — .14 .18 .65 .43 —
+traffic flow change 20 7.14 — — — 2.22 — 10
Against Traffic — 12.16 5.49 — 2.86 8.20 .65 —
+traffic flow change 14.29 34.62 25.00 — — 4.50 — —
Overall Yielding in
2.99 6.20 4.74 .09 .33 2.35 .51 .27
Unmarked Non-Intersection
Unmarked Intersection
With Traffic — — .31 — — — — —
+traffic flow change — — 3.57 — — — — —
Against Traffic — — 5.36 — — — — —
+traffic flow change — — 9.09 — — — — —
Overall Yielding in
— — 3.27 — — — — —
Unmarked Intersection
Grand Percentage .45 .80 1.67 .16 .05 .34 .66 .13
— Indicates no actions of this type were recorded.

73
Vehicle Yielding

Table 15 summarizes the percentage of vehicles that yielded to pedestrians within each crossing
area at each location. As an example, at Location 1, there were three instances of vehicles
yielding to pedestrians with traffic at the unmarked non-intersection. This is 5.17 percent of the
58 total crossings in this area.

The overall percentage of crossings that involved a vehicle yielding to a pedestrian was
compared across locations. The mean percentage of crossings that involved vehicle yielding was
9.77. The Location 7 value of 38.70 percent was 2.15 standard deviations above the mean and
considered an outlier. If the Location 7 value is removed, Location 4 becomes an outlier at
2.05 standard deviations above the mean. If Location 4 is removed from these values, no other
outliers remain. These overall yielding behaviors do not provide much information about where
vehicles are yielding to pedestrians. As a result, yielding behaviors by crossing area are
examined next.

The mean percentage of vehicles that yielded to pedestrians in the marked intersections was
10.21. If all of the locations are compared with one another, the Location 7 value of
39.80 percent was 2.11 standard deviations above the mean and is considered an outlier. If
Location 7 is removed, Location 4 becomes an outlier at 2.05 standard deviations above the
mean. If Location 4 is removed from these values, no other outliers remain.

Next, the mean percentage of vehicles that yielded to pedestrians in unmarked non-intersections
was examined. Overall, the mean percentage of yielding was 1.25. No outliers exist here. In
other words, when looking at these eight locations alone, none are significantly different with
regard to the proportion of drivers who yielded to pedestrians in unmarked non-intersection
areas.

A t-test was performed to determine whether the proportion of vehicle yielding behaviors
differed between the marked intersections and the unmarked non-intersections. No significant
difference between the two locations was found, t(7) = 1.86, p > .05. Next, crossings where the
pedestrian crossed entirely during the walk phase in the marked intersection were compared with
all other vehicle yielding. Once again, there was no significant difference in the percentage of
vehicles that yielded to pedestrians during the walk phase in the marked intersection
(M = 12.62 percent) compared with those vehicles who yielded to vehicles in all other situations
(M = .97 percent), t(7) = 1.81, p > .05.

74
 Table 15. Percentage of vehicles yielding to pedestrians within each crossing area at each
data collection location.
Location
1 2 3 4 5 6 7 8
Marked Intersection
Walk 4.08 7.22 — 29.88 .01 8.74 50.71 .34
+traffic flow change — — — — 28.21 .74 16.57 20
Don’t Walk — — — — — — .62 —
+traffic flow change — — — .99 — — .51 —
Overall Yielding in
3.49 6.95 — 22.34 .12 8.50 39.80 .50
Intersection
Unmarked Non-Intersection
With Traffic 5.17 — — .70 .18 .13 2.13 .38
+traffic flow change — — — — 6.67 — — 10
Against Traffic 4.69 — .61 .27 — — 4.52 2.33
+traffic flow change — — — — — — — —
Overall Yielding in
4.48 — .43 .52 .33 .09 3.07 1.10
Unmarked Non-Intersection
Unmarked Intersection
With Traffic — — .11 — — — — —
+traffic flow change — — 3.57 — — — — —
Against Traffic — — 1.04 — — — — —
+traffic flow change — — — — — — — —
Overall Yielding in
— — .73 — — — — —
Unmarked Intersection
Grand Percentage 3.60 6.35 .35 20.4 .16 7.93 38.70 .64
— Indicates no actions of this type were recorded.

Yielding Comparison

It is important to understand overall yielding behavior. Here pedestrian and vehicle yielding are
compared. First, yielding behaviors that occurred within the marked intersection were explored.
No significant difference in the percentage of pedestrians yielding to vehicles (M = 0.19 percent)
and the percentage of vehicles yielding to pedestrians (M = 10.21 percent), was found,
t(7) = -2.05, p > .05. Similarly, no significant difference in pedestrian yielding (M =
2.19 percent) and vehicle yielding (M = 1.25 percent) in pedestrian crossings occurring in
unmarked non-intersections was found, t(7) = .87, p > .05.

Next, yielding behaviors within the marked intersection that took place entirely during the walk
light phase were examined. A significant difference in the percentage of pedestrians who yielded
to vehicles (M = .18 percent) and the percentage of vehicles who yielded to pedestrians was not
found (M = 12.62 percent), t(7) = -1.95, p > .05. Next, yielding behaviors that took place while
the pedestrian crossed outside the marked intersection or in the marked intersection that was not
entirely during the walk light phase were examined. A paired comparison revealed that there was
no significant difference in the percentage of pedestrians who yielded to vehicles

75
(M = 1.68 percent) and the percentage of vehicles who yielded to pedestrians (M = 0.97 percent),
t(7) = .769, p > .05.

Evasive Pedestrian Actions

Table 16 summarizes the percentage of pedestrian evasive actions in each crossing area at each
of the data collection locations. Each of the three types of evasive actions (running/ accelerated
walking, abrupt stopping, and directional change) was combined to obtain a better overall
perspective on pedestrian evasive actions. As an example, at Location 1, there were 29 instances
of running/accelerated walking, 2 abrupt stops, and 1 directional change in the marked
intersection during the walk phase. This is 3.72 percent of 834 total crossings in this area.

The overall percentage of crossings that involved a pedestrian evasive action was compared
across locations. The mean percentage of crossings that involved a pedestrian evasive action was
4.39. None of the locations had a mean percentage that was more than 2 standard deviations from
this mean. In other words, no outliers existed. These overall values do not provide much
information about where pedestrians are taking evasive actions. As a result, evasive actions by
crossing area are examined next.

The mean percentage of pedestrians who took evasive actions in the marked intersection was
3.44. If all of the locations are compared with one another, none are considered outliers.

Next, the mean percentage of pedestrians who took evasive actions in unmarked non-
intersections was examined. Overall, the mean percentage was 10.17. No outliers exist here. In
other words, when looking at these eight locations, none are significantly different in regard to
the proportion of pedestrians who took evasive actions in unmarked non-intersection areas.

A t-test was performed to determine whether a difference existed in the percentage of pedestrian
evasive actions between the marked intersection and the unmarked non-intersection. Indeed a
significant difference between the two locations was found, t(7) = -2.85, p = .025. Next,
crossings where the pedestrian made the entire crossing during the walk phase in the marked
intersection were compared with all other pedestrian evasive actions. Here, once again, there was
a significant difference in the percentage of pedestrians who took evasive actions during the walk
phase in the marked intersection (M = 2.21 percent) compared with those pedestrians who took
evasive actions in all other situations (M = 11.77 percent), t(7) = -3.51, p = .010.

76
 Table 16. Percentage of pedestrian evasive actions in each crossing area at each data
collection location.
Location
1 2 3 4 5 6 7 8
marked intersection
Walk 3.72 4.96 .22 1.60 1.70 1.83 .24 3.42
+traffic flow change 12.50 25.00 20.0 12.77 74.36 28.15 14.79 80.0
Don’t Walk 28.72 19.75 7.05 4.57 36.54 25.83 8.42 42.86
+traffic flow change 17.50 16.67 — 1.49 40.0 7.48 .51 70.0
Overall Evasive Actions in
6.76 5.56 1.13 2.26 2.18 2.36 2.01 5.29
Intersection
Unmarked Non-Intersection
With Traffic 18.97 4.84 3.85 1.41 2.17 2.30 4.68 13.79
+traffic flow change — — — — 20.00 24.44 — 40.0
Against Traffic 31.25 21.62 9.15 1.92 14.29 11.48 14.84 40.70
+traffic flow change 28.57 26.92 12.50 — — 5.41 — 12.50
Overall Evasive Actions in
24.63 9.18 7.76 1.46 3.32 5.52 8.70 20.82
Unmarked Non-Intersection
Unmarked Intersection
With Traffic — — .94 — — — — —
+traffic flow change — — 7.14 — — — — —
Against Traffic — — 4.08 — — — — —
+traffic flow change — — — — — — — —
Overall Evasive Actions in
— — 2.84 — — — — —
Unmarked Intersection
Grand Percentage 8.92 5.87 2.19 2.19 2.25 2.57 2.21 8.89
— Indicates no actions of this type were recorded.

Evasive Vehicle Actions

Table 17 summarizes the percentage of vehicle evasive actions within each crossing area at each
of the data collection locations. Each of the four types of evasive actions (abrupt braking—first
vehicle, abrupt braking—second vehicle, directional change—first vehicle, and directional
change—second vehicle) were combined to obtain a better overall perspective on vehicle evasive
actions. As an example, at Location 2, there were four instances of abrupt braking by the first
vehicle, two directional changes by the first vehicle, and four directional changes by the second
vehicle in the marked intersection during the walk phase. This is .25 percent of the 4,071 total
crossings at this area.

The overall percentage of crossings that involved a vehicle evasive action was compared across
locations. The mean percentage of crossings that involved a vehicle evasive action was .05. None
of the locations had a mean percentage that was more than 2 standard deviations away from the
mean. In other words, no outliers existed. These overall values do not provide much information
about where vehicles are taking evasive actions. As a result, evasive actions by crossing area are
examined next.

77
The mean percentage of vehicles that took evasive actions in the marked intersection was .04. If
all of the locations are compared with one another, the Location 2 value of .24 percent is
2.27 standard deviations above the mean, making it an outlier. If Location 2 is removed, not
surprisingly, Location 1 (the only other non-zero value) becomes an outlier at 2.27 standard
deviations above the mean.

Next, the mean percentage of vehicles that took evasive actions in unmarked non-intersections
was examined. Overall, the mean percentage was .09. Not surprisingly, Location 1 (the only non-
zero value) was an outlier at 2.47 standard deviations above the mean.

A t-test was performed to determine whether a difference existed in the percentage of vehicle
evasive actions between the marked intersection and the unmarked non-intersection. No
significant difference between the two locations was found, t(7) = -.568, p > .05. Next, crossings
where the pedestrian completed the entire crossing during the walk phase in the marked
intersection were compared with all other vehicle evasive actions. Here, there was no significant
difference in the percentage of vehicles who took evasive actions during the walk phase in the
marked intersection (M = .03 percent) compared with those vehicles who took evasive actions in
all other situations (M = .09 percent), t(7) = -.600, p > .05.

Table 17. Percentage of vehicle evasive actions within each crossing area at each data
collection location.
Location
1 2 3 4 5 6 7 8
Marked Intersection
Walk — .25 — — — — — —
+traffic flow change — — — — — — — —
Don’t Walk 1.06 — — — — — — —
+traffic flow change — — — — — — — —
Overall Evasive Actions in
.10 .24 — — — — — —
Intersection
Unmarked Non-Intersection
With Traffic 1.72 — — — — — — —
+traffic flow change — — — — — — — —
Against Traffic — — — — — — — —
+traffic flow change — — — — — — — —
Overall Evasive Actions in
.75 — — — — — — —
Unmarked Non-Intersection
Unmarked Intersection
With Traffic — — — — — — — —
+traffic flow change — — — — — — — —
Against Traffic — — — — — — — —
+traffic flow change — — — — — — — —
Overall Evasive Actions in
— — — — — — — —
Unmarked Intersection
Grand Percentage .18 .22 — — — — — —
— Indicates no actions of this type were recorded.

78
Evasive Action Comparison

It is important to understand overall evasive action behavior. Here pedestrian and vehicle evasive
actions are compared. First evasive actions that occurred within the marked intersection were
explored. A significant difference in the percentage of pedestrians taking evasive actions
(M = 3.44 percent) and the percentage of vehicles taking evasive actions (M = .04 percent) was
found, t(7) = 4.74, p = .002. Similarly a significant difference in pedestrian evasive actions
(M = 10.17 percent) and vehicle evasive actions (M = 0.09 percent) in pedestrian crossings
occurring in unmarked non-intersections was found, t(7) = 3.54, p = .010.

Next, evasive behaviors within the marked intersection that took place entirely during the walk
light phase were examined. A significant difference in the percentage of pedestrians who took
evasive actions (M = 2.21 percent) and the percentage of vehicles that took evasive actions
(M = .03 percent) was found, t(7) = 3.78, p = .007. Next, evasive actions that took place while
the pedestrian crossed either in the unmarked non-intersection or in the marked intersection at
least partially during the don’t walk phase were examined. Again, there was a significant
difference in the percentage of pedestrians who took evasive actions (M = 11.77 percent) and the
percentage of vehicles that took evasive actions (M = .09 percent), t(7) = 3.83, p = .006.

DISCUSSION

An overarching goal of the present study is to determine which environmental factors influence
where pedestrians cross the roadway. In phase 1 of this study, pedestrian crossing behaviors were
recorded and coded over a 2-week period at eight different locations. It was hoped that these data
would help to identify factors that influence pedestrians to cross at unmarked non-intersection
locations. Furthermore, it was hoped that these data would produce a model that might predict
pedestrian crossings. However, of the total 65,725 crossings, only 4,399 (or 6.7 percent) took
place in an unmarked non-intersection location. In other words, these crossings are generally rare
events. As a result, the number of location areas limits the statistically appropriate modeling
techniques available. These methodologies were unable to successfully model the pedestrian
crossing data from these eight locations.

To attempt to more accurately model and predict pedestrian crossing behavior, data from an
additional 12 locations were collected. These data are described further in phases 2 and 3. Here,
the data and trends for the eight locations from phase 1 are discussed.

Crossing Location

As noted previously, few crossings took place in an unmarked non-intersection location. Overall,
only 6.7 percent of the crossings occurred outside an intersection (marked or unmarked).
Another way to examine these crossings is to explore the distributions of crossings at each
location. Location 8 is considered an outlying value, with 28.2 percent of the crossings occurring
at unmarked non-intersection areas. This warrants further discussion about the characteristics of
Location 8 that might cause this result.

The distance to the next marked crossing for Location 8 was approximately 433 ft. This location
is not considered an outlier (nor are any of the other locations) in the current eight locations,
which have a mean distance of 418.9 ft. The AADT value of 21 is not considered an outlier (nor

79
is the AADT for any of the other locations) in the current eight locations, which have a mean of
20.7. The eight locations also did not significantly vary in the length of the walk phase, the
length of the don’t walk phase, or the width of the crossing. As such, it is not likely that any of
these factors alone caused the above average number of crossings at an unmarked non-
intersection.

Location 8, however, does require pedestrians to travel at a speed of 3.7 ft/s to cross in the
marked intersection entirely during the walk light phase. This rate is greater than the Manual on
Uniform Traffic Control Devices (MUTCD) recommended rate of 3.5 ft/s. Furthermore, the
MUTCD states, “Where pedestrians who walk slower than 3.5 ft/s, or pedestrians who use
wheelchairs, routinely use the crosswalk, a walking speed of less than 3.5 ft/s should be
considered in determining the pedestrian clearance time.”(28) Given that there is an elementary
school approximately two blocks away from this intersection and that there are small suburban
type establishments (including a church, neighborhood market, and library) in the general area, it
is likely that there is substantial pedestrian traffic near this intersection that regularly travels at a
rate of less than 3.5 ft/s.

Although pedestrians were not specifically queried about their crossings, it is possible that
pedestrians may feel rushed while crossing at the marked intersection. This rapid pace required
to cross at the marked intersection during the walk phase may lead pedestrians to feel hurried and
uncomfortable crossing at this location. This intersection is also just outside a traffic circle. This
location prevents pedestrians from being able to see traffic from a distance adequate to determine
whether a vehicle will continue traveling within the traffic circle or exit toward the intersection.
Furthermore, this location reduces pedestrians’ abilities to confidently determine whether the
vehicle will cross their potential path during a crossing that takes place (at least partially) during
the don’t walk phase (a scenario that is likely if the pedestrian is not already waiting at the
intersection and can cross at a rapid pace). This time pressure, combined with a long wait time
for the next walk phase (79 s), may encourage pedestrians to cross outside this marked
intersection.

As noted previously, perceived control of a situation influences pedestrians’ intentions to cross

the roadway. Given the short crossing time and pedestrians’ inability to confidently predict
behavior of traffic exiting from the traffic circle, it is possible that pedestrians may feel that
crossing at unmarked non-intersection areas increases their control of the situation (i.e.,
perceived control). Moving south of the marked intersection to cross the roadway provides
pedestrians the opportunity to see bi-directional vehicle travel from a greater distance and to
potentially find longer gaps in traffic to permit crossing at a more leisurely pace. In addition, the
next marked crossing is not light controlled. As a result, pedestrians must still find an appropriate
gap to cross traffic, and consequently using that marked crossing is not likely to present a greater
level of perceived control. As such, crossing at an unmarked non-intersection may optimize
pedestrians’ perceived control and perceived time efficiency in crossing the roadway.

There were no outliers in terms of the lowest percentage of crossings at unmarked non-
intersection areas. However, Location 7 did have the lowest percentage, only 3.0. This low value
is likely the result of a several factors. First, the distance between the marked intersection of
interest and the next marked crossing is only 145 ft. While this is not an outlier, it is the most
extreme value at 1.7 standard deviations below the mean value of 418.9 ft. Furthermore, there

80
are few trip originators or destinations. Along the east side of the block is a park. However,
shrubbery blocks entrance along this side, and pedestrians need to go to an adjacent side of the
park to obtain entry. In other words, crossing the road between the two marked crossings does
not provide an advantage in gaining entrance to the park. Along the west side of the block, there
is a single large office building. Although it was not specifically recorded, anecdotal evidence
suggests that most pedestrians who crossed at an unmarked non-intersection area were traveling
to or from this office building. Furthermore, along this west side of the block, there are several
large flower planters and concrete benches (that face away from the roadway). These objects do
not prevent pedestrians from crossing in this area, but they do impede a simple and direct travel
path from the sidewalk to the roadway. It is likely that the concrete structures deter pedestrians
from crossing outside the nearby marked crosswalks.

At both Location 7 and Location 4, more than 90 percent of the crossings took place within the
marked intersection. These two locations also had the two highest percentages of crossings that
took place in the marked intersection during the don’t walk light phase (17.9 and 18.7,
respectively, compared with an overall mean of 6.9). As a result, it is important to explore the
factors that might influence these crossings. At both locations, there is a left turn arrow that
guides traffic through the marked intersection that might be difficult to see from the pedestrian’s
perspective.

At Location 7, Pennsylvania Avenue Northwest travels east/west, just north and perpendicular to
the marked intersection of interest. Prior to the walk phase, vehicles traveling west on
Pennsylvania Avenue are given a green arrow to turn south—turning through the marked
intersection. However, pedestrians traveling east at this intersection cannot see that the left
turning vehicles have a green light. Instead these pedestrians can see that cross traffic has a red
light and that the opposite walk signal has just turned from walk to don’t walk. As such,
pedestrians here may incorrectly assume that it is safe to begin crossing, perhaps presuming that
the light is simply in the “delay” time period, and they anticipate that it will turn to the walk
phase shortly after entering the crosswalk. Again, it should be noted that pedestrians were not
asked about their crossings. However, given the unique characteristics of this crossing location,
this is a reasonable and logical explanation of the pedestrian crossing behavior.

At Location 4, 13th Street Northwest travels north/south, just east and perpendicular to the
marked intersection of interest. Just after the conclusion of the walk phase, vehicles traveling
north on 13th Street are given a green arrow to turn west—turning through the marked
intersection. However, pedestrians traveling south at this intersection cannot see that the left
turning vehicles have a green light and may incorrectly assume that the vehicles have a red light
and will stop. Similarly to Location 7, pedestrians here may think that it is safe to cross the
roadway, when indeed this may not be the case.

Location 3 had a large proportion of pedestrian crossings at an unmarked intersection. In fact,

pedestrians were approximately equally likely to cross at the unmarked intersection as the
marked intersection. This finding suggests that pedestrians perceive the unmarked intersection to
be a safe, and acceptable, place to cross the roadway. Although the traffic light phasing was not
set to incorporate pedestrian crossings, pedestrians took advantage of the median separating east
and west traveling traffic. Given that 41 percent of the crossings at the unmarked intersection
involved waiting on the median, it appears that pedestrians were not trapped on the median as

81
previous research has reported. Rather, at this location pedestrians plan their crossing in phases;
crossing a segment and waiting on the median and then completing the crossing. This is a tactic
that presumably increases perceived control.

Beyond the ability to divide the crossing into two portions, environmental factors both encourage
crossing at the unmarked intersection and discourage traveling to the marked intersection.
Pedestrians traveling to/from the north side of Rhode Island Avenue and Marion Street are
required to travel out of the way to cross at the marked intersection. This is a result of the
juxtaposition of the streets. Rhode Island Avenue is a “diagonal” street. This means that
pedestrians need to travel south along Rhode Island in addition to west to reach the marked
intersection, when traveling north. (When traveling south, pedestrians must travel farther south
than desired and then return north while traveling east to reach Marion Street.) Furthermore, the
marked intersection involves three segments (crossings at only the center and main segment were
specifically counted in this study), which, combined, are wider than the single unmarked
intersection. Each segment is divided by a concrete median/island that includes a separate
crosswalk signal. This means that it is possible to begin crossing the street in one light phase, but
the pedestrian may need to wait on a median/island until the next light phase to complete the
crossing. To pedestrians, these two factors combined may outweigh the potential benefits of
crossing during a protected light phase—especially given the relatively rare occurrence of a
vehicle–pedestrian collision. In addition, the unmarked intersection has several factors that
afford a crossing. It is obvious that this is at an intersection—a location where most marked
crossings are found. Further, this location lies at a junction where it is natural to want to travel.
There is Metro station only one block west of the marked intersection, on the north side. As a
result, crossing at the unmarked intersection when traveling to/from the Metro station, along
Marion Street, is likely the most direct and efficient route. Finally, the median in the unmarked
intersection looks like a sidewalk. Pedestrians can clearly see a concrete area on the end of the
median that is approximately the width of a standard sidewalk (e.g., a firm, raised surface that
serves as a barrier from roadway vehicles). As such, this area affords the same things to
pedestrians as a standard sidewalk. It is likely that pedestrians interpret use of this area as they
would any other sidewalk area.

Pedestrian Yielding

In total, 298 of the 65,725 crossings (.45 percent) involved pedestrian yielding. Overall, the
mean percentage of pedestrians who yielded to vehicles was .53 (i.e., the mean of the percentage
of pedestrians who yielded to vehicles at each location).

Overall, Location 3 had a significantly greater percentage of pedestrians who yielded to vehicles.
However, this value is inflated owing to the crossings that occurred in the unmarked intersection
location. As a result, of this unique crossing situation, Location 3 is not discussed further here.

The mean percentage of pedestrians who yielded to vehicles in the marked intersection was .19.
Location 7 was an outlier in the percentage of pedestrians that yielded to vehicles, with
2.42 standard deviations above the mean. This area had a substantial number of vehicles turning
right and passing through the intersection. While the exact circumstances of each yielding
behavior were not recorded, it was noted on several occasions that pedestrians would allow
vehicles to pass to relieve vehicle congestion. In other words, a pedestrian(s) would allow

82
vehicles to complete a right turn so that other traffic traveling straight through might pass. There
were also several instances where a left-turning vehicle yielded to a pedestrian while the vehicle
had a protected turn (i.e., a green left turn arrow), but was left in the intersection when oncoming
vehicular traffic started flowing. In these circumstances, pedestrians often yielded to vehicles
during the walk phase to allow the vehicle to complete the turn and avoid traffic delays.

Overall, significantly more pedestrians yielded to vehicles in the unmarked non-intersections

than in the marked intersections. This suggests that pedestrians who cross outside the marked
intersection are more likely to encounter a situation that requires yielding to avoid collision.

Vehicle Yielding

In total, 9,385 of the 65,725 crossings (14 percent) involved a vehicle yielding to a pedestrian.
Overall, vehicle yielding did not vary by pedestrian crossing location. The mean percentage of
drivers who yielded to pedestrians was 9.77 (i.e., the mean of the percentage of vehicles who
yielded to pedestrians at each location). Overall, Location 7 (38.7 percent) had a significantly
greater percentage of vehicles that yielded to pedestrians. As noted in the previous section, a
large number of vehicles turned through the marked intersection area at Location 7. As a result
of the turning traffic having a green light and the pedestrian signal in the walk phase, vehicles
often waited to complete their turn (i.e., yielded) while pedestrians crossed the roadway. A
similar situation to Location 7 presents itself at Location 4. At Location 4, right turning vehicles
have a green light at the same time that the pedestrian signal is in the walk phase. As a result,
vehicles often waited to complete their turn (i.e., yielded) while pedestrians crossed the roadway.

Yielding Comparison

A surprising result is that there is no significant difference in the percentage of pedestrians who
yielded to vehicles and the percentage of vehicles that yielded to pedestrians during the walk
phase in the marked intersection. One would expect that few pedestrians would yield to vehicles.
This lack of a difference could be the result of several factors.

It is possible (and likely) that pedestrians are aware of the extreme traffic congestion in
Washington, DC. This may lead them to intermittently allow vehicles to complete turns, which
allows following traffic to continue straight and subsequently alleviates minor intersection
congestion. It is also possible that pedestrians simply feel safer in allowing the vehicle to turn
prior to completing their crossing. In other words, if a pedestrian–vehicle collision were to occur,
it is more likely that the pedestrian would be injured than the driver of the vehicle, even at low
speeds. As a result, pedestrians may simply be exhibiting caution.

Evasive Pedestrian Actions

In total, 1,862 of the 65,725 crossings (2.83 percent) involved an evasive pedestrian action.
Overall, the mean percentage of pedestrians who took evasive actions was 4.39 (i.e., the mean of
the percentage of evasive pedestrian actions at each location). The percentage of evasive actions
was fairly consistent across locations and across crossing types. At the present time, the data do
not suggest any clear environmental reason for these evasive pedestrian actions.

83
The percentage of evasive pedestrian actions did, however, vary by pedestrian crossing location.
Pedestrians were more likely to take evasive actions in the unmarked non-intersection areas than
the marked intersection. Likewise, pedestrians were less likely to take evasive actions while
crossing entirely during the walk light phase in the marked crossing than in all other crossings.

Evasive Vehicle Actions

In total, 12 of the 65,725 crossings (.02 percent) involved an evasive vehicle action. Overall, the
mean percentage of vehicles that took evasive actions was .05 (i.e., the mean of the percentage of
evasive vehicle actions at each location). The only evasive vehicle actions took place in
Location 1 and Location 2. This, combined with the small total number of vehicle evasive
actions, does not provide the opportunity to make inferences at this time.

Further, the percentage of vehicle evasive actions did not vary by pedestrian crossing location.
That is, vehicles were equally likely to take an evasive action in the unmarked non-intersection
as in the marked intersection.

Evasive Action Comparison

In all crossing locations, pedestrians were more likely to take an evasive action than vehicles.
This suggests that pedestrians may take a proactive approach to increase their safety by evading
potential collisions with vehicles.

General Discussion

In the present set of data, non-intersection crossings are relatively rare. This makes it difficult to
make predictions about where pedestrians will cross the roadway. Furthermore, of the
65,725 coded crossings, there was only a single close call/near miss. This near miss occurred at
Location 7 during the walk phase in the marked intersection. In this instance, two pedestrians
were about to enter the roadway while at the same time, a vehicle began to make a right turn
(passing through the marked intersection). One person pulled the other person back to the curb.
Although it is not clear from the video, it appears that the vehicle may have nearly collided with
the pedestrian had he or she not been pulled back to the curb. Despite the pedestrian having the
right of way in this case, it is easy to understand why the driver continued with the right turn.
Prior to the pedestrians attempting to cross the roadway, they stood on the curb facing the
opposite crossing direction (i.e., facing north, rather than east). Furthermore, the walk phase had
been initiated for more than 10 s prior to the pedestrians attempting to enter the marked
intersection. These clues could easily lead a driver to interpret the pedestrians as not a potential
hazard.

The subsequent sections discuss data collection and findings from phases 2 and 3.

84
 PEDESTRIAN CROSSING BEHAVIORS: PHASE 2

To more thoroughly assess pedestrian crossing behaviors, seven additional pedestrian crossing
locations in Washington, DC, were selected. Similarly to phase 1, pedestrian crossings were
video recorded and later coded at each of the locations.

DATA COLLECTION LOCATIONS

The same criteria were used to select the second set of cameras as those used in phase 1. Video
feed was recorded from each of the seven cameras from 7:00 a.m. to 5:00 p.m. and from
8:00 p.m. to 11:00 p.m. Eastern Standard Time from February 22, 1012, to March 9, 2012.
The following section describes each of the seven data collection locations, which are labeled in
no particular order.

Location 9

The ninth data collection area is located at the intersection of Connecticut Avenue Northwest and
Florida Avenue Northwest in Washington, DC. Figure 27 is an aerial view of this data collection
area. Connecticut Avenue runs northwest/southeast and is located near the center of the image.
Florida Avenue runs northeast/southwest and is located on the right side of the image. The solid
red rectangle highlights the intersection of Connecticut Avenue Northwest and Florida Avenue
Northwest. The dotted red rectangle highlights the intersection of Connecticut Avenue Northwest
and Leroy Place (west of Connecticut)/T Street (east of Connecticut) Northwest. The DDOT
camera was positioned on the southwest corner of the Connecticut and Florida facing north
(see figure 28). As a result, pedestrians making east/west crossings on Connecticut Avenue
between Florida and Leroy Place/T Street were captured. (Leroy Place meets Connecticut on the
west and T Street meets Connecticut on the east.) The distance from Florida Avenue to Leroy
Place is roughly 551 ft. Approximately 135 ft north of the intersection of Connecticut and
Florida, Bancroft Street meets with the west side of Connecticut. Both right and left turns can be
made from Bancroft onto Connecticut. Figure 28 shows marked crossings are present at both
Florida Avenue Northwest (solid red line) and Leroy Place Northwest (dotted red line). The
AADT in this area is 34 (expressed in thousands and rounded to the nearest 100).(18)

85
Original image: ©2012 Google®; map annotations provided by SAIC.
Figure 27. Photo. Aerial view of Location 9.(29)

The neighborhood around this location is a combination of residential homes and commercial
buildings. The area along the east side of Connecticut Avenue contains a fitness center, a major
drug chain store, a local restaurant, and office space. Along the west side of the same block,
there are non-profit organizations, a church, and a national university satellite location. The area
directly west of this block consists nearly entirely of residential homes.

Figure 28. Photo. Still image captured from the Location 9 camera.

There are three northbound and three southbound lanes. (These lanes continue both north and
south of the relevant data collection block.) In addition, on the northbound side, an additional

86
lane serves as a bus lane near the southern portion. Further north, this lane is available for
metered parking outside rush hours. On the southbound side, an additional lane serves as a right
turn only lane near the intersection with Florida Avenue. Further north, this lane is available for
metered parking outside rush hours.

The marked intersection of Connecticut and Florida Avenues of interest (as highlighted by the
solid rectangular box in figure 28) is 109 ft long (curb to curb). Each of the raised concrete
medians is 5 ft in width. The walk phase is illuminated for 30 s, and the don’t walk phase is
illuminated for 68 s. Figure 29 illustrates what a pedestrian might see as he or she attempts to
cross from the west side of Connecticut Avenue to the east side along Florida Avenue Northwest.

Figure 29. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the west side of Connecticut Avenue Northwest to the east side of the street along
Florida Avenue Northwest.

Location 10

The 10th data collection area is located at the intersection of Connecticut Avenue Northwest and
Van Ness Street Northwest in Washington, DC. Figure 30 is an aerial view of this data collection
area. Connecticut Avenue runs northwest/southeast and is located in the center of the image. Van
Ness Street runs east/west and is located toward the bottom of the image. The solid red rectangle
highlights the intersection of Connecticut Avenue Northwest and Van Ness Street Northwest.

87
The dotted red rectangle highlights the intersection of Connecticut Northwest and Veazey
Terrace. The DDOT camera was positioned on the southwest corner of the Connecticut and Van
Ness facing north (see figure 31). As a result, pedestrians making east/west crossings on
Connecticut Avenue between Van Ness Street and Veazey Terrace were captured. The distance
from Van Ness Street and Veazey Terrace is approximately 361 ft. Figure 30 shows marked
crossings are present at both Van Ness Street Northwest (solid red line) and Veazey Terrace
Northwest (dotted red line). The AADT of this specific block on Connecticut Avenue was not
known. As a result, an average value based on the blocks north and south of the relevant block
was computed for an estimated AADT of 41.8 (expressed in thousands and rounded to the
nearest 100).(18)

Original image: ©2012 Google®; map annotations provided by SAIC.

Figure 30. Photo. Aerial view of Location 10. (30)

The neighborhood around this location is a combination of residential homes and commercial
buildings. The entire west side of Connecticut Avenue is part of the University of the District of
Columbia campus, which consists of mostly green space near the roadway. Along the east side of
Connecticut Avenue in the same block, there is an ATM station, several small local restaurants, a
liquor store, dry cleaning services, chain convenience store, office space, and a service station.
Just north of Veazey Terrace is the Van Ness Metro station, which has entrances on both the east
and west sides of the roadway.

88
 Figure 31. Photo. Still image captured from the Location 10 camera.

During the morning rush hour, there are four southbound lanes and two northbound lanes of
traffic. During the evening rush hour, there are two southbound lanes and four northbound lanes
of traffic. During the remaining times, there are two northbound and two southbound lanes of
traffic, flanked by metered parking on both sides of the roadway. (These lanes continue both
north and south of the relevant data collection block.)

The marked intersection of Connecticut Avenue and Van Ness Street of interest (as highlighted
by the solid rectangular box in figure 31) is 72.5 ft long (curb to curb). The walk phase is
illuminated for 24 s, and the don’t walk phase is illuminated for 75 s. Figure 32 illustrates what a
pedestrian might see as he or she attempts to cross from the west side of Connecticut Avenue to
the east side along Van Ness Street Northwest.

89
 Figure 32. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the west side of Connecticut Avenue Northwest to the east side of the street along Van
Ness Street Northwest.

Location 11

The 11th data collection area is located at the intersection of Georgia Avenue Northwest and
Arkansas Avenue Northwest in Washington, DC. Figure 33 is an aerial view of this data
collection area. Georgia Avenue runs north/south and is located in the center area of the image.
Arkansas Avenue runs northeast/southwest and is located toward the left in the image. The solid
red rectangle highlights the intersection of Georgia Avenue Northwest and Arkansas Avenue
Northwest. The dotted red rectangle highlights the intersection of Georgia Avenue and Farragut
Street. The DDOT camera was positioned on the northeast corner of the Georgia and Arkansas
intersection facing south (see figure 34). As a result, pedestrians making east/west crossings on
Georgia Avenue between Arkansas Avenue and Farragut Street were captured. The distance
from Arkansas Avenue and Farragut Street is approximately 193 ft. Figure 33 shows marked
crossings are present at both Arkansas Avenue Northwest (solid red line) and Farragut Street
Northwest (dotted red line). The intersection of Farragut Street and Georgia Avenue is not light
controlled. The AADT in this area is 22.9 (expressed in thousands and rounded to the
nearest 100).(18)

90
Original image: ©2012 Google®; map annotations provided by SAIC.
Figure 33. Photo. Aerial view of Location 11.(31)

The neighborhood around this location consists of mainly residential homes. The area along the
east side of Georgia Avenue contains an automotive collision center, a service station, and a
convenience store. Along the west side of the same block, there is a park that fills the entire
relevant block.

Figure 34. Photo. Still image captured from the Location 11 camera.

There are two northbound and two southbound lanes of traffic. (These lanes continue both north
and south of the relevant data collection block.) There is a driveway entrance to Georgia Avenue
on the northbound lane. In addition, there is a southbound bus stop near the intersection of
Georgia Avenue and Farragut Street.

91
The marked intersection of Georgia Avenue and Arkansas Avenue of interest (as highlighted by
the solid rectangular box in figure 34) is 68.3 ft long (curb to curb). The walk phase is
illuminated for 19 s, and the don’t walk phase is illuminated for 80 s. Figure 35 illustrates what a
pedestrian might see as he or she attempts to cross from the west side of Georgia Avenue to the
east side along Arkansas Avenue Northwest.

Figure 35. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the west side of Georgia Avenue Northwest to the east side of the street along
Arkansas Avenue Northwest.

Location 12

The 12th data collection area is located at the intersection of Georgia Avenue Northwest and
Irving Street Northwest in Washington, DC. Figure 36 is an aerial view of this data collection
area. Georgia Avenue runs north/south and is located in the center of the image. Irving Street
runs east/west and is located in the lower portion of the image. The solid red rectangle highlights
the intersection of Georgia Avenue Northwest and Irving Street Northwest. The dotted red
rectangle highlights the intersection of Georgia Avenue and Kenyon Street. The DDOT camera
was positioned on the southwest corner of the Georgia and Irving facing north (see figure 37). As
a result, pedestrians making east/west crossings on Georgia Avenue between Irving Street and

92
Kenyon Street were captured. The distance from Irving Street and Kenyon Street is
approximately 277 ft. Figure 36 shows marked crossings are present at both Irving Street
Northwest (solid red line) and Kenyon Street Northwest (dotted red line). The AADT of this
specific block on Georgia Avenue was not known. As a result, an average value based on the
blocks north and south of the relevant block was computed for an estimated AADT of 19.75
(expressed in thousands and rounded to the nearest 100).(18)

Original image: ©2012 Google®; map annotations provided by SAIC.

Figure 36. Photo. Aerial view of Location 12.(32)

The neighborhood around this location is a combination of residential homes and commercial
buildings. The eastern and western sides of Georgia Avenue contain largely the same types of
establishments: barber shops, hair salons, a thrift store, local restaurant, and small neighborhood
markets. Just north of Irving Street, there is a southbound bus stop on Georgia Avenue.

93
 Figure 37. Photo. Still image captured from the Location 12 camera.

There are two northbound and two southbound lanes of traffic, flanked by metered parking on
both sides of the roadway. (These lanes continue both north and south of the relevant data
collection block.)

The marked intersection of Georgia Avenue and Irving Street of interest (as highlighted by the
solid rectangular box in figure 37) is 64.5 ft long (curb to curb). The walk phase is illuminated
for 20 s, and the don’t walk phase is illuminated for 78 s. Figure 38 illustrates what a pedestrian
might see as he or she attempts to cross from the east side of Georgia Avenue to the west side
along Irving Street Northwest.

94
Figure 38. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the east side of Georgia Avenue Northwest to the west side of the street along Irving
Street Northwest.

Location 13

The 13th data collection area is located at the intersection of H Street Northwest and Connecticut
Avenue Northwest. Figure 39 is an aerial view of this data collection area. H Street runs east/
west and is located in the center of the image. Connecticut Avenue runs southeast/northwest and
is located to the left of the image; it creates a t-intersection. The solid red rectangle highlights the
intersection of H Street Northwest and Connecticut Avenue Northwest. The dotted red rectangle
highlights the intersection of H Street and 16th Street. The DDOT camera was positioned on the
northwest corner of the H and Connecticut facing east (see figure 40). As a result, pedestrians
making north/south crossings on H Street between Connecticut Avenue and 16th Street were
captured. The distance from Connecticut Avenue and 16th Street is approximately 316 ft.
Figure 39 shows marked crossings are present at both Connecticut Avenue Northwest (solid red
line) and 16th Street Northwest (dotted red line). The AADT in this area is 15.1 (expressed in
thousands and rounded to the nearest 100).(18)

95
Original image: ©2012 Google®; map annotations provided by SAIC.
Figure 39. Photo. Aerial view of Location 13.(33)

The neighborhood around this location consists of mostly commercial buildings and tourist
attractions. The north side of H Street contains the U.S. Chamber of Commerce and a hotel. The
south side is entirely bordered by Lafayette Square (a park located between the White House and
H Street). Along the south side of H Street there are also several bus and trolley stops.

Figure 40. Photo. Still image captured from the Location 13 camera.

There are four eastbound lanes of traffic (H Street is a one-way street) and a bus lane on the
southern portion of the street during rush hour. During the remaining hours of the day, the

96
northernmost lane is available for metered parking. (These lanes continue east of the relevant
data collection block.)

The marked intersection of H Street and Connecticut Avenue of interest (as highlighted by the
solid rectangular box in figure 40) is 50 ft long (curb to curb). The walk phase is illuminated for
10 s, and the don’t walk phase is illuminated for 89 s. Figure 41 illustrates what a pedestrian
might see as he or she attempts to cross from the north side of H Street to the south side along
Connecticut Avenue Northwest.

Figure 41. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side of H Street Northwest to the south side of the street along Connecticut
Avenue Northwest.

Location 14

The 14th data collection area is located at the intersection of Independence Avenue Southwest
and Washington Avenue Southwest in Washington, DC. Figure 42 is an aerial view of this data
collection area. Independence Avenue runs east/west and is located in the center of the image.
Washington Avenue runs southeast/northwest and is located in the lower portion of the image; it

97
creates a T-intersection with Independence Avenue. The solid red rectangle highlights the
intersection of Independence Avenue Southwest and Washington Avenue Southwest. The dotted
red rectangle highlights the intersection of Independence and 1st. The DDOT camera was
positioned on the northwest corner of Independence and Washington facing east (figure 43). As a
result, pedestrians making north/south crossings on Independence Avenue between Washington
Avenue and 1st Street were captured. The distance from Washington Avenue to 1st Street is
approximately 338 ft. Figure 43 shows marked crossings are present at both Washington Avenue
Southwest (solid red line) and 1st Street Southwest (dotted red line). The AADT in this area is
34.7 (expressed in thousands and rounded to the nearest 100).(18)

Original image: ©2012 Google®; map annotations provided by SAIC.

Figure 42. Photo. Aerial view of Location 14.(34)

The neighborhood around this location consists of mostly government buildings and tourist
attractions. Both the northern and southern portions of Independence Avenue contain the United
States Botanic Garden.

98
 Figure 43. Photo. Still image captured from the Location 14 camera.

There are three eastbound and three westbound lanes of traffic. (These lanes continue east of the
relevant data collection block.)

The marked intersection of Independence Avenue and Washington Avenue of interest (as
highlighted by the solid rectangular box in figure 43) is 68.25 ft long (curb to curb). The walk
phase is illuminated for 30 s, and the don’t walk phase is illuminated for 67 s. Figure 44
illustrates what a pedestrian might see as he or she attempts to cross from the north side of
Independence Avenue to the south side along Washington Avenue Southwest.

99
 Figure 44. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side of Independence Avenue Southwest to the south side of the street along
Washington Avenue Southwest.

Location 15

The 15th data collection area is located at the intersection of Pennsylvania Avenue Southeast and
8th Street Southeast in Washington, DC. Figure 45 is an aerial view of this data collection area.
Pennsylvania Avenue runs southeast/northwest and is located in the center of the image.
Eighth Street runs south/north and is located in right portion of the image. The solid red
rectangle highlights the intersection of Pennsylvania Avenue Southeast and 8th Street Southeast.
The dotted red rectangle highlights the intersection of Pennsylvania and 7th Street. The DDOT
camera was positioned on the northeast corner of Pennsylvania and 8th facing west
(see figure 46). As a result, pedestrians making north/south crossings on Pennsylvania Avenue
between 8th Street and 7th Street were captured. The distance from 8th Street to 7th Street is
approximately 297.5 ft. Figure 45 shows marked crossings are present at both 8th Street
Southeast (solid red line) and 7th Street Southeast (dotted red line). The AADT of this specific
block on Pennsylvania Avenue was not known. As a result, an average value based on the
adjacent blocks of the relevant block was computed for an estimated AADT of 18.3 (expressed
in thousands and rounded to the nearest 100).(18)

100
Original image: ©2012 Google®; map annotations provided by SAIC.
Figure 45. Photo. Aerial view of Location 15.(35)

The neighborhood around this location consists primarily of residential homes with some
commercial buildings. The south side of this block of Pennsylvania Avenue is a park area with
an entrance to the Eastern Market Metro station and a Capital Bikeshare™ station. A wide grass
median divides the southeast bound lanes of traffic and the northwest bound lanes of traffic. A
junior high school spans the north side of Pennsylvania Avenue.

Figure 46. Photo. Still image captured from the Location 15 camera.

101
There are three southeast bound lanes of traffic on Pennsylvania Avenue. (These lanes continue
east of the relevant data collection block.) In addition, there is a fourth bus lane for stops outside
the Metro station. Crossings on the north side of the grass median were not assessed.

The marked intersection of Pennsylvania Avenue and 8th Street of interest (as highlighted by the
solid rectangular box in figure 46) is 45.5 ft long (curb to curb). The walk phase is illuminated
for 39 s, and the don’t walk phase is illuminated for 60 s. Figure 47 illustrates what a pedestrian
might see as he or she attempts to cross from the south side of Pennsylvania Avenue to the north
side along 8th Street Southeast.

Figure 47. Photo. Illustration of what a pedestrian might see as he or she attempts to make
a crossing from the south side of Pennsylvania Avenue Southeast to the north side of the
street along 8th Street Southeast.

DATA COLLECTION VALIDATION

Just as in phase 1, the validity of the camera footage was assessed for the second set of cameras.
Researchers manually scored pedestrian crossing behaviors at Location 10 (Connecticut and Van
Ness Northwest), Location 13 (H and Connecticut Northwest), and Location 15 (Pennsylvania
and 8th Southeast). Researchers recorded the pedestrian crossings and their interactions with
vehicles over three 15-min periods at both locations. The in-vivo recordings were made the
morning of March 1, 2012.

102
At Location 10, there was a 100-percent agreement in the classification of the pedestrian
crossings for both the second and third 15-min segments. However, in the first 15-min session,
there was a single discrepancy; the onsite coding resulted in 25 total pedestrian crossings and the
video coding resulted in 24 crossings. At Location 13, there was a 100-percent agreement in the
classification of the pedestrian crossings for both the second and last 15-min segments. However,
in the first 15-min session, there was a discrepancy; the onsite coding resulted in 81 total
pedestrian crossings and the video coding resulted in 75 crossings. It is likely that this difference
is the result of the large number of people crossing during this short time frame. The video
allowed the researchers to slow the rate of crossing for a theoretically more accurate count in
crossings. At Location 15, there was a 100-percent agreement in the classification of the
pedestrian crossings for both the second and last 15-min segments. However, in the first 15-min
session, there was a discrepancy; the onsite coding resulted in 24 total pedestrian crossings and
the video coding resulted in 22 crossings. Across all three sites, there was a 98.5-percent
agreement in crossing classifications.

VIDEO DATA CODING

Many different types of pedestrian crossings and pedestrian interactions with vehicles in, and
along, the roadway can be recorded. Although it is difficult to code pedestrian crossings in an
exhaustive manner, the current study sought to record enough information to interpret general
crossing behaviors. For each pedestrian crossing, multiple factors were recorded.

Crossing Factors

The same crossing factors as phase 1 were used to classify crossing behaviors in the second set
of cameras.

Dates/Times Coded

Only sub-portions of the vast amount of video recorded data were coded owing to both time and
project scope requirements of the current study. The scope of the present study included only
daytime pedestrian crossings. As such, nighttime data were not examined.

Because there were no significant differences in pedestrian crossing behavior patterns based on
the day of the week or the time of day in phase 1, fewer days and segments of video were coded
in phase 2. Four days were randomly selected to code: February 24, February 27, March 1, and
March 2. On each of these days, samples of crossings were selected throughout the day from
7:00 a.m. to 5:00 p.m., for a total of approximately 1.5 h of coded video per location, per day.
The video feed from Location 11 on March 2 and from Location 13 on February 27 was blurred
and could not be coded.

Weather

Table 18 describes the weather and corresponding sunrise/sunset times for each of the days that
data were coded.

103
 Table 18. Relevant sunrise, sunset, and weather for each of the coded data collection days.
High Low Precipitation
Sunrise Sunset Temperature Temperature General Accumulation
Date (a.m.) (p.m.) (ᵒF) (ᵒF) Weather (inches)
2/24/2012 6:48 5:55 57 46 Rain .28
2/27/2012 6:44 5:59 64 36 Windy —
Mostly
3/01/2012 6:40 6:02 70 45 —
Cloudy
3/02/2012 6:38 6:03 55 41 Rain .40
— Indicates no recorded precipitation.

RESULTS

The follow subsections describe the overall crossing behavior results from phase 2.

Descriptive Data

Table 19 summarizes the crossing behaviors recorded during the coded data collection times. All
values are combined over the seven data collection locations. The following subsections explain
the basic descriptive data for each of the seven locations in further detail.

104
 Table 19. Summary of pedestrian crossings—combined across all seven data collection locations.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 2,002 — 205 5 — — — — — —
+traffic flow
93 — 4 2 — — — — — —
change
Don’t Walk 393 — — 22 - 1 — — — —
+traffic flow
85 1 — 2 1 — — — — —
change
Sum 2,573 1 209 31 1 1 — — — —
Unmarked Non-Intersection
With Traffic 388 1 1 5 — — — — — —
+traffic flow
16 — — 1 — — — — — —
change
Against Traffic 207 18 16 28 1 3 2 — — —
+traffic flow
5 — — 0 — — — — — —
change
Sum 616 19 17 34 1 3 — — — —
Grand Total 3,189 20 226 65 2 4 2 — — —
— Indicates no actions of this type were recorded.

105
Location 9

Total Crossings:

There were a total of 805 pedestrian crossings on Connecticut Avenue Northwest between
Florida Avenue and Leroy Place Northwest. (See table 20 for a summary of these crossings.) Of
these crossings, 717 (or 89.07 percent) occurred within the marked intersection (i.e., marked
crosswalk). These crossings can be broken down into when they occurred during the light
phases. During the walk phase, 625 pedestrians crossed in the marked intersection. An additional
33 crossings occurred in the marked intersection during the don’t walk phase. In addition,
59 crossings occurred in the marked intersection that included a traffic flow change. Of these
crossings, 20 began in the walk phase and ended in the don’t walk phase. An additional 39 began
in the don’t walk phase and concluded in the walk phase.

At Location 9, 88 (or 10.93 percent) of the total 805 crossings took place at the unmarked non-
intersection areas. Of these, 29 took place with traffic, and 53 took place against traffic. A
portion of the pedestrian crossings that took place at an unmarked non-intersection area involved
a traffic flow change. Four of these crossings began with traffic and ended against traffic, and
two began against traffic and ended with traffic.

Yielding:

In total, there were three yielding behaviors (combined across pedestrians and vehicles). In two
instances, pedestrians yielded to vehicles and a single vehicle yielded to a pedestrian at an
unmarked non-intersection location against traffic. No other yielding behaviors were observed.

Evasive Pedestrian Actions:

A total of seven pedestrians took evasive actions at Location 9; all involved running/accelerated
walking. There were three instances of pedestrian running/accelerated walking while in the
marked intersection entirely during the don’t walk phase. In an additional case, a pedestrian
began crossing during the don’t walk phase and completed the crossing during the walk phase.
There were also three instances of pedestrian running/accelerated walking while crossing at and
unmarked non-intersection location.

Evasive Vehicle Actions:

No recorded evasive vehicle actions were recorded at Location 9.

106
 Table 20. Summary of pedestrian crossings at Location 9.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 625 — — — — — — — — —
+traffic flow
20 — — — — — — — — —
change
Don’t Walk 33 — — 3 — — — — — —
+traffic flow
39 — — 1 — — — — — —
change
Sum 717 — — 4 — — — — — —
Unmarked Non-Intersection
With Traffic 29 — — — — — — — — —
+traffic flow
4 — — — — — — — — —
change
Against Traffic 53 2 1 3 — — — — — —
+traffic flow
2 — — — — — — — — —
change
Sum 88 2 1 3 — — — — — —
Grand Total 805 2 1 7 — — — — — —
— Indicates no actions of this type were recorded.

107
Location 10

Total Crossings:

There were a total of 528 pedestrian crossings on Connecticut Avenue Northwest between Van
Ness Street and Veazey Terrace. (See table 21 for a summary of these crossings.) Of these
crossings, 446 (or 84.47 percent) occurred within the marked intersection (i.e., marked
crosswalk). These crossings can be broken down into when they occurred during the light
phases. During the walk phase, 445 pedestrians crossed in the marked intersection. In a single
additional instance, a pedestrian crossed starting in the don’t walk phase but completed the
crossing during the walk light phase.

At Location 10, 82 (15.53 percent) of the total 528 crossings took place in an unmarked non-
intersection area. Of these, 46 took place with traffic, and 34 took place against traffic.

A portion of the pedestrian crossings that took place at an unmarked non-intersection area
involved a traffic flow change. A single crossing began with traffic and ended against traffic, and
a single crossing began against traffic and ended with traffic.

Yielding:

In total, there were 161 yielding behaviors (combined across pedestrians and vehicles). In a
single case, a pedestrian who began crossing during the don’t walk phase, but completed the
crossing during the walk phase, yielded to a vehicle. There were also 10 cases where pedestrians
yielded to vehicles while crossing in an unmarked non-intersection location. One of these
instances occurred with traffic and nine occurred against traffic.

There were 150 cases of vehicles yielding to pedestrians. In 146 of these instances, vehicles
yielded to pedestrians crossing during the walk phase in the marked intersection. The remaining
four instances occurred when pedestrians were crossing at unmarked non-intersection locations
against traffic.

Evasive Pedestrian Actions:

A total of 11 pedestrians took evasive actions at Location 10, and all involved running/
accelerated walking. There were two instances of pedestrian running/accelerated walking while
in the marked intersection during the walk phase. There were also nine instances of pedestrian
running/accelerated walking while crossing at and unmarked non-intersection location.

Evasive Vehicle Actions:

No recorded evasive vehicle actions were recorded at Location 10.

108
 Table 21. Summary of pedestrian crossings at Location 10.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 445 — 146 2 — — — — — —
+traffic flow
— — — — — — — — — —
change
Don’t Walk — — — — — — — — — —
+traffic flow
1 1 — — — — — — — —
change
Sum 446 1 146 2 — — — — — —
Unmarked Non-Intersection
With Traffic 46 1 — 1 — — — — — —
+traffic flow
1 — — — — — — — — —
change
Against Traffic 34 9 4 8 — — — — — —
+traffic flow
1 — — — — — — — — —
change
Sum 82 10 4 9 — — — — — —
Grand Total 528 11 150 11 — — — — — —
— Indicates no actions of this type were recorded.

109
Location 11

Total Crossings:

A total of 17 pedestrian crossings were recorded on Georgia Avenue Northwest between

Arkansas Avenue and Farragut Street Northwest. (See table 22 for a summary of these
crossings.) Of these crossings, 12 (or 70.59 percent) occurred within the marked intersection.
These crossings can be broken down into when they occurred during the light phases. During the
walk phase, there were 10 pedestrian crossings. During the don’t walk phase, there were two
crossings.

There were five total crossings in an unmarked non-intersection area. Of these, three occurred
with traffic and two occurred against traffic.

Yielding:

No yielding behaviors were recorded at Location 11.

Evasive Pedestrian Actions:

No evasive pedestrian actions were recorded at Location 11.

Evasive Vehicle Actions:

No evasive vehicle actions were recorded at Location 11.

110
 Table 22. Summary of pedestrian crossings at Location 11.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 10 — — — — — — — — —
+traffic flow
— — — — — — — — — —
change
Don’t Walk 2 — — — — — — — — —
+traffic flow
— — — — — — — — — —
change
Sum 12 — — — — — — — — —
Unmarked Non-Intersection
With Traffic 3 — — — — — — — — —
+traffic flow
— — — — — — — — — —
change
Against Traffic 2 — — — — — — — — —
+traffic flow
— — — — — — — — — —
change
Sum 5 — — — — — — — — —
Grand Total 17 — — — — — — — — —
— Indicates no actions of this type were recorded.

111
Location 12

Total Crossings:

There were a total of 185 crossings on Georgia Avenue Northwest between Irving Avenue and
Kenyon Street Northwest. (See table 23 for a summary of these crossings.) Of these crossings,
127 (or 68.65 percent) occurred within the marked intersection. These crossings can be broken
down into when they occurred during the light phases. During the walk phase, 120 pedestrians
crossed in the marked intersection. An additional six crossings occurred in the marked
intersection during the don’t walk phase. There was also a single crossing that began during the
walk phase and was completed during the don’t walk phase.

At Location 12, 58 (31.35 percent) of the total 185 crossings took place at the unmarked non-
intersection areas. Of these, 29 took place with traffic, and 27 took place against traffic. Two of
these crossings began with traffic and ended against traffic.

Yielding:

In total, there were 23 yielding behaviors (combined across pedestrians and vehicles) at
Location 12. There were five instances of pedestrians yielding to vehicles while crossing in the
unmarked non-intersection area.

Of the 18 instances of vehicles yielding to pedestrians, 12 occurred while the pedestrian was
crossing during the walk phase in the marked intersection. The remaining seven cases where
vehicles yielded to pedestrians occurred during unmarked non-intersection crossings.

Evasive Pedestrian Actions:

No evasive pedestrian actions were recorded at Location 12.

Evasive Vehicle Actions:

There were two recorded instances of vehicles abruptly braking for pedestrians at Location 12.
Both occurred while pedestrians were crossing at an unmarked non-intersection area.

112
 Table 23. Summary of pedestrian crossings at Location 12.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 120 — 12 2 — — — — — —
+traffic flow
1 — — — — — — — — —
change
Don’t Walk 6 — — 2 — — — — — —
+traffic flow
— — — — — — — — — —
change
Sum 127 — 12 4 — — — — — —
Unmarked Non-Intersection
With Traffic 29 — — — — — — — — —
+traffic flow
2 — — 1 — — — — — —
change
Against Traffic 27 5 6 6 — — 2 — — —
+traffic flow
— — — — — — — — — —
change
Sum 58 5 6 7 — — 2 — — —
Grand Total 185 5 18 11 — — 2 — — —
— Indicates no actions of this type were recorded.

113
Location 13

Total Crossings:

There were a total of 609 crossings on H Street Northwest between Connecticut Avenue and
16th Street Northwest. (See table 24 for a summary of these crossings.) Of these crossings,
554 (or 90.97 percent) occurred within the marked intersection (i.e., marked crosswalk). These
crossings can be broken down into when they occurred during the light phases. During the walk
phase, 173 pedestrians crossed in the marked intersection. An additional 285 crossings occurred
in the marked intersection during the don’t walk phase. In addition, 96 crossings occurred in the
marked intersection that included a traffic flow change. Of these crossings, 66 began in the walk
phase and ended in the don’t walk phase. An additional 30 began in the don’t walk phase and
concluded in the walk phase.

At Location 13, 55 (or 9.03 percent) of the total 609 crossings took place in an unmarked non-
intersection area. Of these, 31 took place with traffic, and 22 took place against traffic. An
additional two of these crossings began with traffic and ended against traffic.

Yielding:

In total, there were 25 yielding behaviors. All of these instances involved vehicles yielding to
pedestrians. Of these, 24 took place while a pedestrian was crossing in the marked intersection
during the walk phase. In a single case, a vehicle yielded to a pedestrian crossing against traffic
in an unmarked non-intersection area.

Evasive Pedestrian Actions:

A total of 20 pedestrians took evasive actions at Location 13. Of these, 18 involved running/
accelerated walking. There was a single instance of pedestrian running/accelerated walking while
in the marked intersection entirely during the walk phase. There were two cases where
pedestrians began crossing during the walk phase but completed their crossing in the don’t walk
phase. In addition, 12 pedestrians ran/accelerated while crossing in the marked intersection
entirely during the don’t walk phase. Three pedestrians ran/accelerated while crossing in an
unmarked non-intersection area.

In a single case, a pedestrian stopped while making a crossing that began during the don’t walk
phase and concluded during the walk phase. There was also a single pedestrian who changed
directions while crossing in the marked intersection during the don’t walk phase.

Evasive Vehicle Actions:

No evasive vehicle actions were recorded at Location 13.

114
 Table 24. Summary of pedestrian crossings at Location 13.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 173 — 24 1 — — — — — —
+traffic flow
66 — — 2 — — — — — —
change
Don’t Walk 285 — — 12 — 1 — — — —
+traffic flow
30 — — — 1 — — — — —
change
Sum 554 — 24 15 1 1 — — — —
Unmarked Non-Intersection
With Traffic 31 — — — — — — — — —
+traffic flow
2 — — — — — — — — —
change
Against Traffic 22 — 1 3 — — — — — —
+traffic flow
— — — — — — — — — —
change
Sum 55 — 1 3 — — — — — —
Grand Total 609 — 25 18 1 1 — — — —
— Indicates no actions of this type were recorded.

115
Location 14

Total Crossings:

There were a total of 205 pedestrian crossings on Independence Avenue Southwest between
Washington Avenue and 1st Street Southwest. (See table 25 for a summary of these crossings.)
Of these crossings, 184 (or 89.76 percent) occurred within the marked intersection (i.e., marked
crosswalk). These crossings can be broken down into when they occurred during the light
phases. During the walk phase, 165 pedestrians crossed in the marked intersection. An additional
10 crossings occurred in the marked intersection during the don’t walk phase. In addition, nine
crossings occurred in the marked intersection that included a traffic flow change. Of these
crossings, three began in the walk phase and ended in the don’t walk phase. An additional six
began in the don’t walk phase and concluded in the “walk phase.”

At Location 14, 21 (or 10.24 percent) of the total 205 crossings took place at the unmarked non-
intersection areas. Of these, 13 took place with traffic and 7 took place against traffic. One of
these crossings began against traffic and ended with traffic.

Yielding:

In total, there were 27 yielding behaviors (combined across pedestrians and vehicles). In
two instances, pedestrians yielded to vehicles at an unmarked non-intersection location against
traffic.

There were 23 instances of vehicles yielding to pedestrians crossing in the marked intersection
during the walk phase and a single case where a pedestrian began crossing during the walk phase
but completed the crossing during the don’t walk phase. There was also a single case where a
vehicle yielded to a pedestrian crossing in the unmarked non-intersection area.

Evasive Pedestrian Actions:

A total of four pedestrians took evasive actions at Location 14; all involved running/accelerated
walking. There were two instances of pedestrian running/accelerated walking while in the
marked intersection entirely during the don’t walk phase. There were also two instances of
pedestrian running/accelerated walking while crossing at an unmarked non-intersection location.

Evasive Vehicle Actions:

No evasive vehicle actions were recorded at Location 14.

116
 Table 25. Summary of pedestrian crossings at Location 14.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 165 — 23 — — — — — — —
+traffic flow
3 — 1 — — — — — — —
change
Don’t Walk 10 — — 2 — — — — — —
+traffic flow
6 — — — — — — — — —
change
Sum 184 — 24 2 — — — — — —
Unmarked Non-Intersection
With Traffic 13 — 1 2 — — — — — —
+traffic flow
— — — — — — — — — —
change
Against Traffic 7 2 — — — — — — — —
+traffic flow
1 — — — — — — — — —
change
Sum 21 2 1 2 — — — — — —
Grand Total 205 2 25 4 — — — — — —
— Indicates no actions of this type were recorded.

117
Location 15

Total Crossings:

There were a total of 840 pedestrian crossings on Pennsylvania Avenue Southeast between
8th Street and 7th Street Southeast. (See table 26 for a summary of these crossings.) Of these
crossings, 533 (or 63.45 percent) occurred within the marked intersection (i.e., marked
crosswalk). These crossings can be broken down into when they occurred during the light
phases. During the walk phase, 464 pedestrians crossed in the marked intersection. An additional
57 crossings occurred in the marked intersection during the don’t walk phase. There were also
11 crossings that occurred in the marked intersection that included a traffic flow change. Of these
crossings, three began in the walk phase and ended in the don’t walk phase. An additional nine
began in the don’t walk phase and concluded in the walk phase.

At Location 15, 307 (or 36.55 percent) of the total 840 crossings took place at the unmarked non-
intersection areas. Of these, 237 took place with traffic, and 62 took place against traffic. A
portion of the pedestrian crossings that took place at an unmarked non-intersection area involved
a traffic flow change. Seven of these crossings began with traffic and ended against traffic, and
one began against traffic and ended with traffic.

Yielding:

In total, there were seven yielding behaviors. There were three instances of vehicles yielding to a
pedestrian crossing in the marked intersection that began in the walk phase but concluded in the
don’t walk phase. There were also four cases of vehicles yielding to pedestrians crossing in an
unmarked non-intersection area. No other yielding behaviors were observed.

Evasive Pedestrian Actions:

A total of 18 pedestrians took evasive actions at Location 15. There were three instances of
pedestrian running/accelerated walking while in the marked intersection entirely during the don’t
walk phase. In an additional case, a pedestrian began crossing during the don’t walk phase and
completed the crossing during the walk phase. There were also 10 instances of pedestrian
running/accelerated walking while crossing at and unmarked non-intersection location.

In a single case, a pedestrian stopped for a vehicle while crossing at an unmarked non-
intersection area. In three instances, pedestrians changed directions while at an unmarked non-
intersection area.

Evasive Vehicle Actions:

No evasive vehicle actions were recorded at Location 15.

118
 Table 26. Summary of pedestrian crossings at Location 15.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 464 — — — — — — — — —
+traffic flow
3 — 3 — — — — — — —
change
Don’t Walk 57 — — 3 — — — — — —
+traffic flow
9 — — 1 — — — — — —
change
Sum 533 — 3 4 — — — — — —
Unmarked Non-Intersection
With Traffic 237 — — 2 — — — — — —
+traffic flow
7 — — — — — — — — —
change
Against Traffic 62 — 4 8 1 3 — — — —
+traffic flow
1 — — — — — — — — —
change
Sum 307 — 4 10 1 3 — — — —
Grand Total 840 — 7 14 1 3 — — — —
— Indicates no actions of this type were recorded.

119
Predictive Model

Based on the results from the first eight cameras (and similarly low percentages of non-
intersection crossings), another predictive modeling attempt was not made for cameras 9 through
15. Instead, modeling was completed after additional data were collected in phase 3.

Factor-Specific Analyses

This results section evaluates each of the recorded crossing variables in detail.

Crossing Location

Table 27 summarizes the percentage of pedestrians, by location, who crossed at marked

intersections and unmarked non-intersections. The percentage of each type of crossing is
presented at each location. For example, at Location 9, 77.64 percent of the total crossings took
place in the marked intersection during the walk phase. When all locations are combined, the
percentage of pedestrians who crossed at the marked intersection is 80.68. The mean percentage
of pedestrians crossing at each of the seven locations is 79.57. None of the locations are outliers.

Table 27. Percentage of pedestrians at each crossing area in each data collection location.
Location
9 10 11 12 13 14 15
Marked Intersection
Walk 77.64 84.28 58.82 64.87 28.41 80.49 55.24
+traffic flow change 2.48 — — .54 10.84 1.46 .36
Don’t Walk 4.10 — 11.75 3.24 46.80 4.88 6.79
+traffic flow change 4.84 .19 — — 4.93 2.93 1.07
Overall in Intersection 89.07 84.47 70.59 68.65 90.97 89.76 63.45
Unmarked Non-Intersection
With Traffic 3.60 8.71 17.65 15.68 5.09 6.34 28.21
+traffic flow change .50 .19 — 1.08 .33 — .83
Against Traffic 6.58 6.44 11.76 14.60 3.61 3.41 7.38
+traffic flow change .25 .19 — — — .49 .12
Overall in Unmarked Non-
10.93 15.53 29.41 31.35 9.03 10.24 36.55
Intersection
— Indicates no actions of this type were recorded.

Pedestrian Yielding

Table 28 summarizes the percentage of pedestrians who yielded to vehicles within each crossing
type. As an example, at Location 9, of the 53 total crossings made against traffic in the unmarked
non-intersection area, 2 pedestrians yielded to a vehicle. This is the equivalent of 3.77 percent of
the total crossings in this specific crossing type. The mean percentage of pedestrians yielding to
vehicles across each of the seven locations is .86. If all of the locations are compared with one
another, none are outliers.

120
The mean percentage of pedestrians who yielded to vehicles in the marked intersection was .03.
If all of the locations are compared with one another, Location 10, the only non-zero value, is not
surprisingly considered an outlier and is 2.27 standard deviations above the mean.

Next, the mean percentage of pedestrians who yielded to vehicles in unmarked non-intersections
was examined. Overall, the mean percentage of yielding was 4.66. No outliers exist here. In
other words, when looking at these seven locations alone, none are significantly different with
regard to the proportion of pedestrians who yielded to vehicles in unmarked non-intersection
areas.

A t-test was performed to determine whether the proportion of pedestrian yielding behaviors
differed between the marked intersection and the unmarked non-intersection. No significant
difference between the two locations was found, t(6) = -2.34, p > .05.

Crossings where the pedestrian completed the entire crossing during the walk phase in the
marked intersection were compared with all other pedestrian yielding. Here, there were no
instances of pedestrians yielding while crossing in the marked intersection entirely during the
walk light phase. There was no significant difference in the percentage of pedestrians who
yielded to vehicles during the walk phase in the marked intersection (M = 0 percent) compared
with those pedestrians who yielded to vehicles in all other situations (M = 3.87 percent),
t(6) = -2.00, p > .05.

Table 28. Percentage of pedestrians yielding to vehicles in each crossing area at each data
collection location.
Location
9 10 11 12 13 14 15
Marked Intersection
Walk — — — — — — —
+traffic flow change — — — — — — —
Don’t Walk — — — — — — —
+traffic flow change — 100 — — — — —
Overall in Intersection — .22 — — — — —
Unmarked Non-Intersection
With Traffic — 2.17 — — — — —
+traffic flow change — — — — — — —
Against Traffic 3.77 26.47 — 18.52 — 28.57 —
+traffic flow change — — — — — — —
Overall in Unmarked Non-
2.27 12.2 — 8.62 — 9.52 —
Intersection
Grand Percentage .25 2.08 — 2.70 — .98 —
— Indicates no actions of this type were recorded.

Vehicle Yielding

Table 29 summarizes the percentage of vehicles that yielded to pedestrians within each crossing
area at each location. As an example, at Location 14, there were 23 instances of vehicles yielding

121
to pedestrians crossing in the walk phase in the marked intersection. This is 13.94 percent of the
165 total walk phase crossings at this location.

The overall percentage of crossings that involved a vehicle yielding to a pedestrian was
compared across locations. The mean percentage of crossings that involved vehicle yielding was
9.71. The Location 10 value of 28.41 percent was 2.00 standard deviations above the mean and
considered an outlier. If the Location 10 value is removed, no other outliers remain. These
overall yielding behaviors do not provide much information about where vehicles are yielding to
pedestrians. As a result, yielding behaviors by crossing area are examined next.

The mean percentage of vehicles that yielded to pedestrians in the marked intersection was 8.59.
If all of the locations are compared with one another, the Location 10 value of 32.74 percent was
2.05 standard deviations above the mean and is considered an outlier. If Location 10 is removed,
no other outliers remain.

Next, the mean percentage of drivers that yielded to pedestrians in unmarked non-intersections
was examined. Overall, the mean percentage of yielding was 3.46. No outliers exist here. In
other words, when looking at these seven locations alone, none are significantly different in the
proportion of vehicles that yielded to pedestrians in unmarked non-intersection areas.

A t-test was performed to determine whether a difference existed in the percentage of vehicle
yielding behaviors between the marked intersections and the unmarked non-intersections. No
significant difference between the two locations was found, t(6) = 1.28, p > .05. Next, crossings
where the pedestrian crossed entirely during the walk phase in the marked intersection were
compared with all other vehicle yielding. Here, there was no significant difference in the
percentage of vehicles that yielded to vehicles during the walk phase in the marked intersection
(M = 10.09 percent) compared with those vehicles that yielded to vehicles in all other situations
(M = 3.10 percent), t(6) = 1.70, p > .05.

122
 Table 29. Percentage of vehicles yielding to pedestrians within each crossing area at each
data collection location.
Location
9 10 11 12 13 14 15
Marked Intersection
Walk — 32.81 — 10 13.87 13.94 —
+traffic flow change — — — — — 33.33 100
Don’t Walk — — — — — — —
+traffic flow change — — — — — — —
Overall in Intersection — 32.74 — 9.45 4.33 13.04 .56
Unmarked Non-Intersection
With Traffic — — — — — 7.69 —
+traffic flow change — — — — — — —
Against Traffic 1.89 11.76 — 22.22 4.55 — 6.45
+traffic flow change — — — — — — —
Overall in Unmarked Non-
1.14 4.88 — 10.34 1.82 4.76 1.30
Intersection
Grand Percentage .12 28.41 — 9.73 4.11 12.20 .83
— Indicates no actions of this type were recorded.

Yielding Comparison

It is important to understand overall yielding behavior. Here pedestrian and vehicle yielding are
compared. First, yielding behaviors in the marked intersection were examined. No significant
difference between pedestrian (M = .03 percent) and vehicle (M = 8.59 percent) yielding was
found, t(6) = -1.93, p > .05. Similarly, when examining yielding behavior in the unmarked non-
intersection, no significant difference between pedestrian (M = 4.66) and vehicle
(M = .89 percent) yielding was found, t(6) = -1.93, p > .05.

Next, yielding behaviors within the marked intersection that took place entirely during the walk
light phase were examined. A t-test revealed that there was no significant difference in the
percentage of pedestrians who yielded to vehicles (M = 0 percent) and the percentage of vehicles
who yielded to pedestrians (M = 10.09 percent), t(6) = -2.24, p > .05. Next, yielding behaviors
that took place while the pedestrian crossed either in the unmarked non-intersection or at least
partially during the don’t walk phase in the marked intersection were examined. There was no
significant difference in the percentage of pedestrians who yielded to vehicles (M = 3.86 percent)
and the percentage of vehicles who yielded to pedestrians (M = 3.10 percent), t(6) = 0.579,
p > .05.

Evasive Pedestrian Actions

Table 30 summarizes the percentage of pedestrian evasive actions within each crossing area at
each of the data collection locations. Each of the three types of evasive actions (running/
accelerated walking, abrupt stopping, and directional change) was combined to obtain a better
overall perspective on pedestrian evasive actions.

123
The overall percentages of crossings that involved pedestrian evasive actions were compared
across locations. The mean percentage of crossings that involved a pedestrian evasive action was
2.32. None of the locations had a mean percentage that was more than 2 standard deviations
away from this mean. In other words, no outliers existed. These overall values do not provide
much information about where pedestrians are making evasive actions. As a result, evasive
actions by crossing area are examined next.

The mean percentage of pedestrians who took evasive actions in the marked intersection was
1.30. If all of the locations are compared with one another, none are considered outliers.

Next, the mean percentage of pedestrians who took evasive actions in unmarked non-
intersections was examined. Overall, the mean percentage was 6.57. No outliers exist here. In
other words, when looking at these eight locations, none are significantly different with regard to
the proportion of pedestrians who took evasive actions in unmarked non-intersection areas.

A t-test was performed to determine whether a difference existed in the percentage of pedestrian
evasive actions between the marked intersection and the unmarked non-intersection. A
significant difference between the two locations was found, t(6) = -3.51, p = .013. Next,
crossings where the pedestrian crossed entirely during the walk phase in the marked intersection
were compared with all other pedestrian evasive actions. Here, there was again a significant
difference in the percentage of pedestrians who took evasive actions during the walk phase in the
marked intersection (M = .39 percent) compared with those pedestrians who took evasive actions
in all other situations (M = 6.82 percent), t(6) = -3.82, p = .009.

Table 30. Percentage of pedestrian evasive actions within each crossing area at each data
collection location.
Location
9 10 11 12 13 14 15
Marked Intersection
Walk — .45 — 1.67 .58 — —
+traffic flow change — — — — 3.03 — —
Don’t Walk 9.09 — — 33.33 4.56 20 5.26
+traffic flow change 2.56 — — — 33.33 — 11.11
Overall in Intersection .56 .45 — 3.15 3.07 1.09 .75
Unmarked Non-Intersection
With Traffic — 2.17 — — — 15.38 .84
+traffic flow change — — — 50.00 — — —
Against Traffic 5.66 23.53 — 22.22 13.64 — 19.35
+traffic flow change — — — — — — —
Overall in Unmarked Non-
3.41 10.98 — 12.07 5.45 9.52 4.56
Intersection
Grand Percentage .87 2.08 — 5.95 3.28 1.95 2.14
— Indicates no actions of this type were recorded.

124
Evasive Vehicle Actions

Table 31 summarizes the percentage of vehicle evasive actions within each crossing area at each
of the data collection locations. Each of the four types of evasive actions (abrupt braking—first
vehicle, abrupt braking—second vehicle, directional change—first vehicle, and directional
change—second vehicle) were combined to obtain a better overall perspective on vehicle evasive
actions.

Location 12 was the only location where evasive actions were observed. As a result, no further
analyses were completed.

Table 31. Percentage of vehicle evasive actions within each crossing area at each data
collection location.
Location
9 10 11 12 13 14 15
Marked Intersection
Walk — — — — — — —
+traffic flow change — — — — — — —
Don’t Walk — — — — — — —
+traffic flow change — — — — — — —
Overall in Intersection — — — — — — —
Unmarked Non-Intersection
With Traffic — — — — — — —
+traffic flow change — — — — — — —
Against Traffic — — — 7.41 — — —
+traffic flow change — — — — — — —
Overall in Unmarked Non-
— — — 3.45 — — —
Intersection
Grand Percentage — — — 1.08 — — —
— Indicates no actions of this type were recorded.

Evasive Action Comparison

Because no vehicles took evasive actions, a comparison between vehicle and pedestrian evasive
actions was not made.

DISCUSSION

An overarching goal of the present study is to determine which environmental factors influence
where pedestrians cross the roadway. Pedestrian crossing behaviors were recorded and coded
over a 2-week period at seven different locations. It was hoped that these data would help to
identify factors that influence pedestrians to cross at unmarked non-intersection locations. As
was discovered in phase 1 of this study, the number of non-intersection crossings is quite low. As
a result, modeling and predicting pedestrian behavior is difficult using statistical methodologies.
To more accurately describe pedestrian crossing behaviors, additional data are collected in

125
phase 3 and described in more detail later. At present, the data and trends for the seven locations
from phase 2 are discussed.

Crossing Location

As in phase 1, the majority of pedestrian crossed at marked intersection locations. Although none
of the locations in phase 2 were considered outlying values in terms of crossing locations, three
locations (11, 12, and 15) did have greater unmarked non-intersection crossing percentages than
those in phase 1. These locations are discussed in more detail here.

Overall, Location 11 had very few crossings. However, this does not imply that there was no
pedestrian traffic in this area. Rather, pedestrians did not have reason to cross the roadway at the
specific location. While not specifically coded, it did appear that the majority of people crossing
in the vicinity of Location 11 crossed at the far non-signalized marked crossing. This may have
been the result of the increased perceived control of crossing when desired. This crossing is also
quite close to a bus stop waiting area. Because of the small sample size (17 crossings), a small
number of incidents can artificially influence the percentage of different types of crossings. As a
result, Location 11 is excluded from further discussions.

Overall, Location 12 is quite representative of many city blocks in many cities. However, several
factors, when combined, may contribute to a high number of unmarked non-intersection
crossings. The first of these factors is that there are a large number of businesses on both sides of
the street with parking located on both sides of the street. The result is that people frequently
park on one side of the street and walk to a business on the opposite side. Further, the businesses
along this block are mixed. This can result in people walking from their workplace to the other
side of the street to obtain food. Although in many cases pedestrians may opt to walk to the
marked crossing to cross the roadway, this portion of Georgia Avenue Northwest is relatively
narrow, with a relatively low AADT. The low AADT provides gaps to cross the roadway that
many pedestrians find acceptable. It is possible that pedestrians perceive that they will have more
time to safely cross the roadway in these gaps than the forced 20-s crossing time (a crossing rate
of 3.23 ft/s) at the marked intersection.

At Location 15, 36.55 percent of the pedestrians crossed at an unmarked non-intersection

location. These crossings took place primarily in one specific area between the two marked
crossings on Pennsylvania Avenue Southeast. On the south side of the relevant block, there are
two primary pedestrian traffic originators/destinations: a Metro (subway) station and a bus stop
that services multiple bus lines. On the north side of the relevant block, there is a large
residential neighborhood. As such, many people travel between their homes and the public
transportation stations/stops. Figure 45 shows there is a short section of roadway on the north
side of Pennsylvania Avenue Southeast that connects westbound traffic to an intersecting street.
This creates another vehicle intersection just west of 8th Street. While this is likely designed to
alleviate vehicular traffic, it creates a “shortcut” for pedestrians. Pedestrians can see a clear
pathway between the Metro station/bus stop areas to the residential area. In using this path to
cross, pedestrians are able to make half of their crossing and wait on the grassy area that divides
the north and south portions of Pennsylvania Avenue before completing the crossing, as needed.
This is a tactic that would likely increase pedestrians’ perceived control of their crossing. As

126
figure 45 shows, pedestrians do indeed use this pathway regularly, so much so that the grass in
the median area has been worn away.

At Location 13, most of the crossings did occur at the marked intersection (89.76 percent).
However, more than half (62.57 percent) of these took place at least partially during the don’t
walk phase. This extreme difference warrants discussion. There are several factors in this area
that would draw people to cross at the marked intersection rather than an unmarked non-
intersection. As figure 39 shows, there is a large park just to the south of the relevant block of
H Street. This park provides the opportunity to walk “diagonally” through several blocks without
encountering vehicle traffic. There are, however, small barriers around the park that inhibit
entrance to the park outside pre-specified areas. As a result of pedestrians’ perceived advantages
of walking through the park, they may also perceive an advantage to cross at the marked
intersection to gain easier access to the park. This however, does not explain why pedestrians
may have been persuaded to cross outside the walk phase.

As was mentioned in phase 1, it appears that the rate at which pedestrians must cross during walk
phase may influence when and where pedestrians decide to cross the roadway. At Location 13,
pedestrians must cross at a rate of 5 ft/s, which greatly exceeds the MUTCD recommendation of
3.5 to 3.7 ft/s.(28) At a rate of 5 ft/s (10 s to cross the roadway) it is possible that many pedestrians
are simply not able to complete the crossing during the allotted time and are forced to complete
the crossing during the don’t walk phase. In addition, other environmental factors in this area
may influence pedestrians to cross during the don’t walk phase. There are two general
expectation violations involved in this crossing. Because of southbound traffic on Connecticut
Avenue turning left, pedestrians crossing west of the marked crossing of interest are allotted
more time to cross the roadway. In other words, pedestrians waiting to cross in the marked
intersection where crossings were recorded, were likely able to see pedestrians making a similar
crossing in the adjacent crosswalk. Furthermore, the pedestrians may have been able to see the
adjacent walk sign and perhaps presumed that there was an error or other problem with their
respective walk sign and entered the roadway. In this case, pedestrians’ expectations to be able to
cross at the same time as the pedestrians in the adjacent crossing were violated. Pedestrians
crossing at this marked intersection can also view the time countdown on the perpendicular
crossing light (i.e., the amount of time left for pedestrians travelling perpendicularly to cross).
They might expect their own crossing light to change to walk shortly thereafter and preemptively
begin crossing the roadway. This behavior may be particularly risky for pedestrians crossing
from north to south because they cannot see the vehicles making left turns (and may not make
appropriate evasive maneuvers). As mentioned in phase 1, it is also possible that pedestrians may
not be able to see that vehicles have a turning arrow and inappropriately begin their crossings.
This certainly could have been the case for pedestrians crossing from south to north in this area.

Pedestrian Yielding

In total, 20 of the 3,189 crossings (.62 percent) involved pedestrian yielding. Overall the mean
percentage of pedestrians who yielded to vehicles was .86 (i.e., the mean of the percentage of
pedestrians who yielded to vehicles at each location). There were no significant differences
between locations when looking at total percentage of pedestrian yielding behaviors.

127
Only at Location 10 were there instances of pedestrians crossing in the marked intersection.
However, the value of .22 percent of the total crossings within the marked intersection is within
the range of observed values in phase 1. As a result, these crossings are not discussed further
here. There were also no significant differences in the percentage of pedestrians who yielded to
vehicles in unmarked non-intersections. All values were similar to those found in phase 1.

Vehicle Yielding

In total, 226 of the 3,189 crossings (7.09 percent) involved a vehicle yielding to a pedestrian.
Overall, the mean percentage of vehicles that yielded to pedestrians was 9.71 (i.e., the mean of
the percentage of vehicles that yielded to pedestrians at each location). This value is quite similar
to phase 1.

Overall, Location 10 (28.41 percent) had a significantly greater percentage of vehicles that
yielded to pedestrians. This high percentage is likely the result of turning vehicle traffic. Further
exploration of this location reveals that the majority of these occurred in the marked intersection
during the walk light phase. Vehicles originating from Van Ness Street have a green light to turn
north on to Connecticut Avenue while pedestrians have a walk light phase. This then, would
commonly result in vehicles yielding to pedestrians in the marked crosswalk.

There were also no significant differences in the percentage of drivers who yielded to pedestrians
based on the area where pedestrians crossed the roadway.

Yielding Comparison

In phase 2, there were no significant differences in the percentage of pedestrians who yielded to
vehicles and the percentage of vehicles that yielded to pedestrians.

Evasive Pedestrian Actions

In total, 71 of the 3,189 crossings (2.23 percent) involved an evasive pedestrian action. This
value was similar to that for phase 1 (2.83 percent). The mean percentage of pedestrians who
took evasive actions was 2.32 (i.e., the mean of the percentage of evasive pedestrian actions at
each location). The percentage of evasive actions was fairly consistent across locations and
across crossing types. At the present time, the data do not suggest any clear environmental
reason for these evasive pedestrian actions.

Evasive Vehicle Actions

In total, 2 of the 3,189 crossings (.06 percent) involved an evasive vehicle action. This value is
similar to that for phase 1 (.02 percent). The only evasive vehicle actions took place in
Location 12. This, combined with the small total number of vehicle evasive actions, does not
provide the opportunity to make inferences at this time.

General Discussion

In the present set of data, non-intersection crossings are relatively rare. This makes it difficult to
make predictions about where pedestrians will cross the roadway. In this second phase, there

128
were no recorded close calls or near misses. In phase 3, researchers recorded crossings at five
sites in vivo. The results of these observations are discussed in the next section.

129
 PEDESTRIAN CROSSING BEHAVIORS: PHASE 3

To more thoroughly assess pedestrian crossing behaviors, five additional pedestrian crossing
locations in the Washington, DC, metropolitan area (including locations in Maryland and
Virginia) were selected. Unlike in phases 1 and 2, data were coded in vivo by two researchers.

DATA COLLECTION LOCATIONS

Nearly the same criteria were used to select the locations used in the third set of data collection
as those used in phase 1 and phase 2. The only difference was that researchers were placed in
locations that could record pedestrian crossings and video cameras were not used. Data were
collected in the late morning/early afternoons on several different days. Data for Locations 16
and 17 were collected on November 21, 2012. Data for Locations 18 and 19 were collected on
December 3, 2012, and data for Location 20 were collected on December 4, 2012.

The following section describes each of the five data collection locations.

Location 16

The 16th data collection area is located at the intersection of King and Washington Streets in
Alexandria, VA. Figure 48 is an aerial view of this data collection area. North Washington Street
runs north/south and is located near the center of the image. King Street runs east/west and is
also located toward the center of the image. The solid red rectangle highlights the intersection of
King Street and North Washington Street. The dotted red rectangle highlights the intersection of
North Washington Street and Cameron Street. Researchers recorded pedestrians crossing
east/west on North Washington Street between King Street and Cameron Street. Researchers
were positioned on the northwest corner of King and North Washington Streets. One researcher
faced north and recorded pedestrians making east/west crossings on North Washington Street
between King Street and Cameron Streets. Another researcher faced east and recorded
pedestrians crossing east/west in the northernmost marked intersection of North Washington and
King Streets. The distance from King Street to Cameron Street is approximately 361 ft.
Figure 48 shows marked crossings are present at both King Street (solid red line) and Cameron
Street (dotted red line). The AADT in this area is 28 (expressed in thousands and rounded to the
nearest 100).(36)

131
Original image: ©2012 Google®; map annotations provided by SAIC.
Figure 48. Photo. Aerial view of Location 16.(37)

The neighborhood around this location is a combination of residential homes and commercial
buildings. Along both the eastern and western sides of North Washington Street, there are
clothing stores, restaurants, and other small shops.

There are two northbound and two southbound lanes of North Washington Street. (These lanes
continue both north and south of the relevant data collection block.) In addition, on the
southbound side, an additional lane serves as a bus lane near the southern portion. Farther north,
metered parking is available outside rush hours. On the northbound side, an additional lane
serves metered parking outside rush hours.

The marked intersection of King and North Washington Streets of interest (as highlighted by the
solid rectangular box in figure 48) is 75 ft long (curb to curb). A median separates the
northbound and southbound traffic. This median is 7.5 ft wide and contains some flower boxes
and trees (see figure 49). The walk phase is illuminated for 58 s, and the don’t walk phase is
illuminated for 73 s.

132
 Figure 49. Photo. Illustration of some of the foliage present in the median of North
Washington Street.

Figure 50 illustrates what a pedestrian might see as he or she attempts to cross from the east side
of North Washington Street to the west side along King Street.

133
Figure 50. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the east side to the west side of North Washington Street along King Street.

Location 17

The 17th data collection area is located at the intersection of King and Washington Streets in
Alexandria, VA. Figure 51 is an aerial view of this data collection area. North Washington Street
runs north/south and is located just to the left of the center of the image. King Street runs
east/west and is located toward the center of the image. The solid red rectangle highlights the
intersection of King Street and North Washington Street. The dotted red rectangle highlights the
intersection of King Street and North Saint Asaph Street. Researchers recorded pedestrians
crossing north/south on King Street between North Washington Street and North Saint Asaph
Street. Researchers were positioned on the southeast corner of King and North Washington
Streets. One researcher faced north and recorded pedestrians making north/south crossings in the
easternmost marked intersection of North Washington and King Streets. Another researcher
faced east and recorded north/south crossings on King Street between North Washington Street
and North Saint Asaph Street. The distance from North Washington Street to North Saint Asaph
Street is approximately 266.5 ft. Figure 51 shows marked crossings are present at both North
Washington Street (solid red line) and North Saint Asaph Street (dotted red line) along King
Street. The AADT in this area is 8.1 (expressed in thousands and rounded to the nearest 100).(36)

134
Original image: ©2012 Google®; map annotations provided by SAIC.
Figure 51. Photo. Aerial view of Location 17.(37)

The neighborhood around this location is a combination of residential homes and commercial
buildings. Along both the northern and southern sides of King Street are clothing stores,
restaurants, and other small shops. This is primarily a shopping and commercial area.

There is one eastbound and one westbound lane of traffic on the relevant block of King Street.
Parking is available on both sides of the street. Near the intersection of King Street and North
Washington Street, the westbound traffic is provided with a right turn only lane to turn north.
Just before the beginning of this lane, there is a bus stop.

The marked intersection of King and North Washington Streets of interest (as highlighted by the
solid rectangular box in figure 51) is 43.25 ft long (curb to curb). The walk phase is illuminated
for 64 s, and the don’t walk phase is illuminated for 63 s.

Figure 52 illustrates what a pedestrian might see as he or she attempts to cross from the south
side of King Street to the north side along North Washington Street.

135
Figure 52. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the south side to the north side of King Street along North Washington Street.

Location 18

The 18th data collection area is located at the intersection of 14th Street Northwest and Irving
Street Northwest in Washington, DC. Figure 53 is an aerial view of this data collection area.
Irving Street Northwest runs east/west and is located in the center of the image. Fourteenth Street
runs north/south and is located in the right half of the image. The solid red rectangle highlights
the intersection of 14th and Irving Streets Northwest. The dotted red rectangle highlights the
intersection of Irving Street and Hiatt Place Northwest in Washington, DC. Researchers recorded
pedestrians crossing north/south on Irving Street between 14th Street and Hiatt Place.
Researchers were positioned on the northwest corner of Irving and 14th Streets. One researcher
faced south and recorded pedestrians making north/south crossings in the westernmost marked
intersection of Irving and 14th Streets. Another researcher faced west and recorded north/south
crossings on Irving Street between 14th Street and Hiatt Place.

The distance from 14th Street Northwest to Hiatt Place is approximately 511.25 ft. Figure 53
shows marked crossings at both 14th Street (solid red line) and Hiatt Place (dotted red line) along
Irving Street. The AADT value for this specific block as not recorded. The value here was
estimated based on a block to the east and to the west of the relevant roadway section. The
estimated AADT in this area is 7.85 (expressed in thousands and rounded to the nearest 100).(18)

136
Original image: ©2012 Google®; map annotations provided by SAIC.
Figure 53. Photo. Aerial view of location 18.(38)

The neighborhood around this location is a combination of residential homes and commercial
buildings. Along both the northern and southern sides of Irving Street Northwest are various
restaurants and commercial businesses. Both the southwest and northeast corners of the relevant
intersection contain entrances to the Metro (subway). The block northwest of this area also
houses a school.

The relevant block of Irving Street Northwest is a one-way street that contains two lanes of
eastbound traffic. Near the intersection of 14th Street, an additional lane is provided for left turns
only. There is no parking along either side of the street. There is a bus stop near the intersection
of Irving and 14th Streets on the south side of Irving.

The marked crosswalk at the relevant intersection of 14th Street and Irving Street Northwest (as
highlighted by the solid rectangular box in figure 53) is 30 ft (curb to curb). The narrower
portion of the street (that does not include the additional turn lane) is 20 ft (curb to curb). The
walk phase is illuminated for 60 s, and the don’t walk phase is illuminated for 40 s.

Figure 54 illustrates what a pedestrian might see as he or she attempts to cross from the north
side of Irving Street Northwest to the south side along 14th Street Northwest.

137
Figure 54. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the north side to the south side of Irving Street Northwest along 14th Street
Northwest in Washington DC.

Location 19

The 19th data collection area is located at the intersection of Wisconsin Avenue and Willard
Avenue in Chevy Chase, Maryland. Figure 55 is an aerial view of this data collection area.
Wisconsin Avenue runs southeast/northwest and is located in the center of the image. Willard
Avenue runs east/west and is located in the upper portion of the image. The solid red rectangle
highlights the intersection of Willard Avenue and Wisconsin Avenue. The dotted red rectangle
highlights the intersection of Wisconsin Avenue and Western Avenue. Researchers recorded
pedestrians crossing northeast/southwest on Wisconsin Avenue between Willard Avenue and
Western Avenue. Researchers were positioned on the southwest corner of Wisconsin and Willard
Avenues. One researcher faced east and recorded pedestrians making east/west crossings in the
southernmost marked intersection of Willard and Wisconsin Avenues. Another researcher faced
southeast and recorded northeast/southwest crossings on Wisconsin Avenue between Willard
Avenue and Western Avenue.

138
The distance from Willard Avenue to Western Avenue is approximately 342 ft. Figure 55 shows
marked crossings are present at both Willard Avenue (solid red line) and Western Avenue
(dotted red line) along Wisconsin Avenue. The AADT value was not available for this specific
block. As a result the value was estimated based on the block of Wisconsin Avenue directly
south of the relevant roadway segment. The estimated AADT in this area is 23.2 (expressed in
thousands and rounded to the nearest 100).(18)

Original image: ©2012 Google®; map annotations provided by SAIC.

Figure 55. Photo. Aerial view of location 19.(39)

The neighborhood around this location is a combination of residential homes and commercial
buildings. Along the southwest side of Wisconsin Avenue, there are many smaller retail shops.
There are also several planters and elevated brick obstacles located between the roadway and
sidewalk that can be seen in figure 56.

139
Figure 56. Photo. View of the planters located between the southbound lanes of traffic and
the adjacent sidewalk along Wisconsin Avenue.

Along the northeast side of Wisconsin Avenue, there is a small convenience store, and entrance
to a commercial building. This area, however, primarily serves as an entrance and awning to a
large bus stop location. There are Metro (subway) entrances on both sides of Wisconsin Avenue
near Western Avenue.

The relevant block of Wisconsin Avenue contains two lanes of southbound traffic. Near the
intersection of Willard Avenue, two additional lanes act as dedicated left and right turn only
lanes. There are also three lanes of northbound traffic. Near the intersection with Willard
Avenue, an additional lane serves as a dedicated left turn only lane. There is a single bus stop
along this side of the roadway near the intersection of Willard Avenue.

The marked crosswalk at the relevant intersection of Wisconsin Avenue and Willard Avenue (as
highlighted by the solid rectangular box in figure 55) is 80 ft (curb to curb). A median divides
north and southbound traffic. It is approximately 5 ft wide, and as figure 57 shows, it is more
difficult to traverse in some areas because of increased elevation and foliage. The walk phase is
illuminated for 30 s, and the don’t walk phase is illuminated for 70 s.

140
 Figure 57. Photo. Illustration of some of the foliage and elevation present in the median of
Wisconsin Avenue.

Figure 58 illustrates what a pedestrian might see as he or she attempts to cross from the north
side of Irving Street Northwest to the South side along 14th Street Northwest.

141
Figure 58. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the west side to the east side of Wisconsin Avenue along Willard Avenue.

Location 20

The 20th data collection area is located at the intersection of Clarendon Boulevard and North
Edgewood Street in Arlington, VA. Figure 59 is an aerial view of this data collection area.
Clarendon Boulevard runs northeast/southwest and is located toward the center of the image.
North Edgewood runs north/south and is located in the center of the image. The solid red
rectangle highlights the intersection of Clarendon Boulevard and North Edgewood Street. The
dotted red rectangle highlights the intersection of Clarendon Boulevard and North Fillmore
Street in Arlington, VA. Researchers recorded pedestrians crossing north/south Along Clarendon
Boulevard between North Fillmore Street and North Edgewood Street. Researchers were
positioned on the southwest corner of Clarendon Boulevard and North Edgewood Street. One
researcher faced north and recorded pedestrians making north/south crossings in the westernmost
marked intersection of Clarendon Boulevard and North Edgewood Street. Another researcher
faced west and recorded north/south non-intersection crossings on Clarendon Boulevard between
Edgewood Street and North Fillmore Street.

The distance from North Edgewood Street to North Fillmore Street is approximately 339.5 ft.
Figure 59 shows marked crossings are present at both North Edgewood Street (solid red line) and

142
North Fillmore Street (dotted red line) along Clarendon Boulevard. The AADT in this area is 13
(expressed in thousands and rounded to the nearest 100).(36)

Original image: ©2012 Google®; map annotations provided by SAIC.

Figure 59. Photo. Aerial view of Location 20.(40)

The neighborhood around this location is a combination of residential homes and commercial
buildings. Along both the northern and southern sides of Clarendon Boulevard are various
restaurants and commercial businesses. The relevant block of Clarendon Boulevard is a one-way
street that contains two lanes of eastbound traffic. Metered parking is available on both sides of
the roadway.

The marked crosswalk at the relevant intersection of Clarendon Boulevard and North Edgewood
Street (as highlighted by the solid rectangular box in figure 59) is 44.5 ft (curb to curb). The walk
phase is illuminated for 20 s, and the don’t walk phase is illuminated for 63 s. Figure 60
illustrates what a pedestrian might see as he or she attempts to cross from the south side of
Clarendon Boulevard to the north side along North Edgewood Street.

143
Figure 60. Photo. Illustration of what a pedestrian might see as he or she attempts to cross
from the south side to the north side of Clarendon Boulevard along North Edgewood
Street.

DATA CODING

For each pedestrian crossing, multiple factors were recorded. The following subsections describe
each of these factors and how data were coded.

Crossing Factors

The same crossing factors as phase 1 and phase 2 were used to classify crossing behaviors in the
third data collection phase.

Dates/Times Coded

Unlike in phase 1 and phase 2, all data were collected through direct (not video recorded)
observation. Data at Location 16 and Location 17 were collected on November, 21, 2012. Data at
Location 18 and Location 19 were collected on December 3, 2012. Data at Location 20 were
recorded on December 4, 2012. All data in phase 3 were collected midday.

144
Weather

Table 32 describes the weather and corresponding sunrise/sunset times for each of the days that
data were coded.

Table 32. Relevant sunrise, sunset, and weather for each of the coded data collection days.
High Low Precipitation
Sunrise Sunset Temperature Temperature General Accumulation
Date (a.m.) (p.m.) (ᵒF) (ᵒF) Weather (inches)
Mostly
11/21/2012 6:07 4:51 57 40 —
Sunny
Mostly
12/03/2012 7:10 4:46 70 46 —
Sunny
12/04/2012 7:11 4:46 69 49 Cloudy —
— Indicates no recorded precipitation.

RESULTS

The follow subsections describe the overall crossing behavior results from phase 3.

Descriptive Data

Table 33 summarizes the crossing behaviors recorded during the coded data collection times. All
values are combined over the five data collection locations. The following subsections explain
the basic descriptive data for each of the five locations in further detail.

145
 Table 33. Summary of pedestrian crossings—combined across data collection locations 16–20.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 1,082 12 82 — — — 1 — — —
+traffic flow
72 — — 8 — — — — — —
change
Don’t Walk 74 — — 3 — 2 — — — —
+traffic flow
33 — — 1 — — — — — —
change
Sum 1,261 12 82 12 — 2 1 — — —
Unmarked Non-Intersection
With Traffic 140 3 14 4 — — — — — —
+traffic flow
1 — — — — — — — — —
change
Against Traffic 56 3 7 2 1 — — — — —
+traffic flow
6 — — — — — — — — —
change
Sum 203 6 21 6 1 — — — — —
Grand Total 1,464 18 103 18 1 2 1 — — —
— Indicates no actions of this type were recorded.

146
Location 16

Total Crossings:

There were a total of 280 pedestrian crossings on North Washington Street between King Street
and Cameron Street. (See table 34 for a summary of these crossings.) Of these crossings, 262 (or
93.57 percent) occurred within the marked intersection (i.e., marked crosswalk). These crossings
can be broken down into when they occurred during the light phases. During the walk phase,
232 pedestrians crossed in the marked intersection. An additional eight crossings occurred in the
marked intersection during the don’t walk phase. In addition, 22 crossings occurred in the
marked intersection that included a traffic flow change. Of these crossings, 16 began in the walk
phase and ended in the don’t walk phase. An additional six began in the don’t walk phase and
concluded in the walk phase.

At Location 16, 18 (or 6.43 percent) of the total 280 crossings took place at the unmarked non-
intersection areas. Of these, 14 took place with traffic, and 1 took place against traffic. A portion
of the pedestrian crossings that took place at an unmarked non-intersection area involved a traffic
flow change. All three of these crossings began against traffic and ended with traffic.

Yielding:

In total, there were 58 yielding behaviors (combined across pedestrians and vehicles). There was
a single instance of a pedestrian yielding to a vehicle in an unmarked non-intersection area. In
the remaining 57 instances of yielding, the vehicle yielded to the pedestrian. In 55 of these
instances, pedestrians were walking in the marked intersection during the walk phase. In the
remaining two cases, the vehicle yielded to a pedestrian in the unmarked non-intersection. No
other yielding behaviors were observed.

Evasive Pedestrian Actions:

A total of 10 pedestrians took evasive actions at Location 16, and all involved running/
accelerated walking. In eight cases, pedestrians began crossing during the walk phase and
completed the crossing in the don’t walk phase. There was a single instance of pedestrian
running/accelerated walking while in the marked intersection entirely during the don’t walk
phase. In an additional case, a pedestrian began crossing during the don’t walk phase and
completed the crossing during the walk phase. There were also three instances of pedestrian
running/accelerated walking while crossing at and unmarked non-intersection location.

Evasive Vehicle Actions:

There were two recorded evasive vehicle actions at Location 16. In both cases, the vehicle
abruptly stopped for a pedestrian crossing in the marked intersection during the don’t walk
phase.

147
 Table 34. Summary of pedestrian crossings at Location 16.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 232 — 55 — — — — — — —
+traffic flow
16 — — 8 — — — — — —
change
Don’t Walk 8 — — 1 — 2 — — — —
+traffic flow
6 — — 1 — — — — — —
change
Sum 262 — 55 10 — 2 — — — —
Unmarked Non-Intersection
With Traffic 14 1 2 — — — — — — —
+traffic flow
— — — — — — — — — —
change
Against Traffic 1 — — — — — — — — —
+traffic flow
3 — — — — — — — — —
change
Sum 18 1 2 — — — — — — —
Grand Total 280 1 57 10 — 2 — — — —
— Indicates no actions of this type were recorded.

148
Location 17

Total Crossings:

There were a total of 225 pedestrian crossings on King Street between North Washington Street
and Saint Asaph Street. (See table 35 for a summary of these crossings.) Of these crossings, 184
(or 81.77 percent) occurred within the marked intersection (i.e., marked crosswalk). These
crossings can be broken down into when they occurred during the light phases. During the walk
phase, 165 pedestrians crossed in the marked intersection. An additional 13 crossings occurred in
the marked intersection during the don’t walk phase. In addition, six crossings occurred in the
marked intersection that included a traffic flow change. Of these crossings, four began in the
walk phase and ended in the don’t walk phase. An additional two began in the don’t walk phase
and concluded in the walk phase.

At Location 17, 41 (or 18.22 percent) of the total 225 crossings took place at the unmarked non-
intersection areas. Of these, 24 took place with traffic, and 17 took place against traffic.

Yielding:

In total, there were 38 yielding behaviors (combined across pedestrians and vehicles). In
15 cases, pedestrians yielded to vehicles. Twelve of these occurred while the pedestrian was in
the marked intersection during the walk phase. In the remaining three instances, pedestrians
yielded to vehicles while crossing in the unmarked non-intersection area.

There were 23 instances of vehicles yielding to pedestrians. In 16 cases, pedestrians were in the
marked intersection during the walk phase. In the remaining seven instances, vehicles yielded to
pedestrians crossing in an unmarked non-intersection.

Evasive Pedestrian Actions:

Only a single evasive pedestrian action was recorded at Location 17. During the don’t walk
phase of in the marked intersection, a pedestrian was recorded as running/accelerated walking.

Evasive Vehicle Actions:

There was a single recorded evasive vehicle action at Location 17. In this case, the vehicle
abruptly stopped for a pedestrian crossing in the marked intersection during the walk phase.

149
 Table 35. Summary of pedestrian crossings at Location 17.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 165 12 16 — — — 1 — — —
+traffic flow
4 — — — — — — — — —
change
Don’t Walk 13 — — 1 — 2 — — — —
+traffic flow
2 — — — — — — — — —
change
Sum 184 12 16 1 — — 1 — — —
Unmarked Non-Intersection
With Traffic 24 1 4 — — — — — — —
+traffic flow
— — — — — — — — — —
change
Against Traffic 17 2 3 — — — — — — —
+traffic flow
— — — — — — — — — —
change
Sum 41 3 7 — — — — — — —
Grand Total 225 15 23 1 — — 1 — — —
— Indicates no actions of this type were recorded.

150
Location 18

Total Crossings:

There were a total of 786 pedestrian crossings on Irving Street Northwest between 14th Street
Northwest and Hiatt Place Northwest. (See table 36 for a summary of these crossings.) Of these
crossings, 661 (or 84.10 percent) occurred within the marked intersection (i.e., marked
crosswalk). These crossings can be broken down into when they occurred during the light
phases. During the walk phase, 574 pedestrians crossed in the marked intersection. An additional
33 crossings occurred in the marked intersection during the don’t walk phase. In addition,
54 crossings occurred in the marked intersection that included a traffic flow change. Of these
crossings, 43 began in the walk phase and ended in the don’t walk phase. An additional 11 began
in the don’t walk phase and concluded in the walk phase.

At Location 18, 125 (or 15.90 percent) of the total 786 crossings took place at the unmarked non-
intersection areas. Of these, 98 took place with traffic, and 26 took place against traffic. In a
single case, a pedestrian began crossing in the unmarked non-intersection with traffic and ended
against traffic.

Yielding:

In total, 12 yielding behaviors (combined across pedestrians and vehicles) were recorded at
Location 18. In all 12 instances, vehicles yielded to pedestrians in an unmarked non-intersection
area.

Evasive Pedestrian Actions:

A total of four pedestrians took evasive actions at Location 18; all involved running/accelerated
walking. In all four instances, pedestrians were recorded as running/ accelerated walking while
crossing at and unmarked non-intersection location.

Evasive Vehicle Actions:

No evasive vehicle actions were recorded at Location 18.

151
 Table 36. Summary of pedestrian crossings at Location 18.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 574 — — — — — — — — —
+traffic flow
43 — — — — — — — — —
change
Don’t Walk 33 — — — — — — — — —
+traffic flow
11 — — — — — — — — —
change
Sum 661 — — — — — — — — —
Unmarked Non-Intersection
With Traffic 98 — 8 3 — — — — — —
+traffic flow
1 — — — — — — — — —
change
Against Traffic 26 — 4 1 — — — — — —
+traffic flow
— — — — — — — — — —
change
Sum 125 — 12 4 — — — — — —
Grand Total 786 — 12 4 — — — — — —
— Indicates no actions of this type were recorded.

152
Location 19

Total Crossings:

There were a total of 84 pedestrian crossings on Wisconsin Avenue between Willard Street and
Western Avenue. (See table 37 for a summary of these crossings.) Of these crossings, 81 (or
96.43 percent) occurred within the marked intersection (i.e., marked crosswalk). These crossings
can be broken down into when they occurred during the light phases. During the walk phase,
73 pedestrians crossed in the marked intersection. In addition, eight crossings occurred in the
marked intersection that included a traffic flow change. Of these crossings, seven began in the
walk phase and ended in the don’t walk phase. An additional one crossing began in the don’t
walk phase and concluded in the walk phase.

At Location 19, 3 (or 3.57 percent) of the total 84 crossings took place at the unmarked non-
intersection areas. Of these, one took place with traffic, and one took place against traffic. In a
single case, a pedestrian began crossing in the unmarked non-intersection against traffic and
ended with traffic.

Yielding:

In total, there were 11 yielding behaviors (combined across pedestrians and vehicles) recorded at
Location 19. In all 11 instances, vehicles yielded to pedestrians in the marked intersection during
the walk phase.

Evasive Pedestrian Actions:

No evasive pedestrian actions were recorded at Location 19.

Evasive Vehicle Actions:

No evasive vehicle actions were recorded at Location 19.

153
 Table 37. Summary of pedestrian crossings at Location 19.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 73 — 11 — — — — — — —
+traffic flow
7 — — — — — — — — —
change
Don’t Walk — — — — — — — — — —
+traffic flow
1 — — — — — — — — —
change
Sum 81 — 11 — — — — — — —
Unmarked Non-Intersection
With Traffic 1 — — — — — — — — —
+traffic flow
— — — — — — — — — —
change
Against 1 — — — — — — — — —
+traffic flow
1 — — — — — — — — —
change
Sum 3 — — — — — — — — —
Grand Total 84 — 11 — — — — — — —
— Indicates no actions of this type were recorded.

154
Location 20

Total Crossings:

There were a total of 89 pedestrian crossings on Clarendon Boulevard between North Edgewood
Street and North Fillmore Street. (See table 38 for a summary of these crossings.) Of these
crossings, 73 (or 82.02 percent) occurred within the marked intersection (i.e., marked crosswalk).
These crossings can be broken down into when they occurred during the light phases. During the
walk phase, 38 pedestrians crossed in the marked intersection. An additional 20 crossings
occurred in the marked intersection during the don’t walk phase. In addition, 15 crossings
occurred in the marked intersection that included a traffic flow change. Of these crossings, two
began in the walk phase and ended in the don’t walk phase. An additional 13 crossings began in
the don’t walk phase and concluded in the “walk phase.”

At Location 20, 16 (or 17.98 percent) of the total 89 crossings took place at the unmarked non-
intersection areas. Of these, 3 took place with traffic, and 11 took place against traffic. In two
instances, pedestrians began crossing in the unmarked non-intersection against traffic and ended
with traffic.

Yielding:

In total, there were two yielding behaviors (combined across pedestrians and vehicles) recorded
at Location 20. In both instances, pedestrians yielded to vehicles while crossing in the unmarked
non-intersection area.

Evasive Pedestrian Actions:

A total of four evasive pedestrian actions were recorded at Location 20. In a single instance, a
pedestrian abruptly stopped while crossing in an unmarked non-intersection area. The remaining
three evasive actions involved running/accelerated walking. In one instance, the pedestrian
accelerated in the marked intersection during the don’t walk phase. In the remaining two cases,
pedestrians were seen running/accelerated walking in unmarked non-intersection areas.

Evasive Vehicle Actions:

No evasive vehicle actions were recorded at Location 20.

155
 Table 38. Summary of pedestrian crossings at Location 20.
Yielding Evasive Pedestrian Actions Evasive Vehicle Actions
Abrupt Abrupt Directional Directional
Running/ Braking Braking Change Change
Total Accelerated Abrupt Directional (1st (2nd (1st (2nd
Crossings Pedestrian Vehicle Walking Stopping Change Vehicle) Vehicle) Vehicle) Vehicle)
Intersection
Walk 38 — — — — — — — — —
+traffic flow
2 — — — — — — — — —
change
Don’t Walk 20 — — 1 — — — — — —
+traffic flow
13 — — — — — — — — —
change
Sum 73 — — 1 — — — — — —
Unmarked Non-Intersection
With Traffic 3 1 — 1 — — — — — —
+traffic flow
— — — — — — — — — —
change
Against Traffic 11 1 — 1 1 — — — — —
+traffic flow
2 — — — — — — — — —
change
Sum 16 2 — 2 1 — — — — —
Grand Total 89 2 — 3 1 — — — — —
— Indicates no actions of this type were recorded.

156
Predictive Model

It was hoped that the results of this study might be used to develop a model that would predict
whether a pedestrian would cross at a marked intersection or an unmarked non-intersection based
on the features of the environment. However, based on the results of the first eight cameras in
phase 1 (and similarly low percentages of non-intersection crossings), another predictive
modeling attempt was not made for locations 16–20. Instead, a model is presented in the next
section that included data from all three data collection phases.

Factor-Specific Analyses

This section evaluates each of the recorded crossing variables in detail.

Crossing Location

Table 39 summarizes the percentage of pedestrians, by location, who crossed at a marked

intersection and unmarked non-intersection. The percentage of each type of crossing is presented
at each location. For example, at Location 19, 86.90 percent of the total crossings took place in
the marked intersection during the walk phase. When the raw values at all locations are
combined, the percentage of pedestrians who crossed at the marked intersection is 86.13. The
mean percentage of pedestrians crossing at each of the five locations is 87.98 in the marked
intersection. None of the locations are outliers.

Table 39. Percentage of pedestrians at each crossing area in each data collection location.
Location
16 17 18 19 20
Marked Intersection
Walk 82.86 73.33 73.03 86.90 42.70
+traffic flow change 5.71 1.78 5.47 8.33 2.25
Don’t Walk 2.86 5.78 4.20 — 22.47
+traffic flow change 2.14 .89 1.40 1.19 14.61
Overall in Intersection 93.57 81.78 84.10 96.43 82.02
Unmarked Non-Intersection
With Traffic 5.00 1.67 12.47 1.19 3.37
+traffic flow change — — .13 — —
Against Traffic .36 7.56 3.31 1.19 12.36
+traffic flow change 1.07 — — 1.19 2.25
Overall in Unmarked Non-Intersection 6.43 18.22 15.90 3.57 17.98
— Indicates no actions of this type were recorded.

Pedestrian Yielding

Table 40 summarizes the percentage of pedestrians who yielded to vehicles within each crossing
type. As an example, at Location 17, of the 165 total crossings made in the marked intersection
during the walk phase, 12 pedestrians yielded to vehicles. This is the equivalent of 7.27 percent
of the total crossings in this specific crossing type. The mean percentage of pedestrians yielding

157
to vehicles across each of the five locations is 1.86. If all of the locations are compared with one
another, none are outliers.

The mean percentage of pedestrians who yielded to vehicles in the marked intersection was 1.30.
If all of the locations are compared with one another, none are outliers (even though Location 17
is the only non-zero value).

Next, the mean percentage of pedestrians who yielded to vehicles in unmarked non-intersections
was examined. Overall, the mean percentage of yielding was 5.08. No outliers exist here. In
other words, when looking at these five locations alone, none are significantly different in the
proportion of pedestrians who yielded to vehicles in unmarked non-intersection areas.

A t-test was performed to determine whether the proportion of pedestrian yielding behaviors
differed between the marked intersection and the unmarked non-intersection. No significant
difference between the two locations was found, t(4) = -1.56, p > .05. Pedestrian yielding that
took place in crossings that occurred within the marked intersection entirely during the walk
phase (M = 1.45 percent) were compared with all other crossings (M = 2.20 percent). No
significant difference was found, t(4) = -.71, p > .05.

Table 40. Percentage of pedestrians yielding to vehicles in each crossing area at each data
collection location.
Location
16 17 18 19 20
Marked Intersection
Walk — 7.27 — — —
+traffic flow change — — — — —
Don’t Walk — — — — —
+traffic flow change — — — — —
Overall in Intersection — 6.52 — — —
Unmarked Non-Intersection
With Traffic 7.14 4.17 — — 33.33
+traffic flow change — — — — —
Against Traffic — 11.76 — — 9.09
+traffic flow change — — — — —
Overall in Unmarked Non-Intersection 5.56 7.32 — — 12.50
Grand Percentage .36 6.67 — — 2.25
— Indicates no actions of this type were recorded.

Vehicle Yielding

Table 41 summarizes the percentage of vehicles that yielded to pedestrians within each crossing
area at each location. As an example, at Location 16, there were 23 instances of vehicles yielding
to pedestrians crossing in the walk phase in the marked intersection. This is 23.71 percent of the
232 total walk phase crossings in this location.

158
The overall percentages of crossings that involved a vehicle yielding to a pedestrian were
compared across locations. The mean percentage of crossings that involved vehicle yielding was
9.04. None of the locations were considered to be outliers. These overall yielding behaviors do
not provide much information about where vehicles are yielding to pedestrians. As a result,
yielding behaviors by crossing area are examined next.

The mean percentage of vehicles that yielded to pedestrians in the marked intersection was 5.57.
None of the locations were considered to be outliers.

Next, the mean percentage of vehicles that yielded to pedestrians in unmarked non-intersections
was examined. Overall, the mean percentage of yielding was 7.56. No outliers exist here. In
other words, when looking at these five locations alone, none are significantly different for the
proportion of drivers who yielded to pedestrians in unmarked non-intersection areas.

A t-test was performed to determine whether a difference existed in the percentage of vehicle
yielding behaviors between the marked intersections and the unmarked non-intersections. No
significant difference between the two locations was found, t(4) = .234, p > .05. Uniquely, in
these five locations, the only instances of vehicles yielding to pedestrians in the marked
intersection occurred during the walk light phase. Vehicle yielding to pedestrians that crossed
within the marked intersection entirely during the walk phase (M = 9.69 percent) were compared
with all other crossings (M = 4.30 percent). No significant difference was found, t(4) =
1.08, p > .05.

Table 41. Percentage of vehicles yielding to pedestrians in each crossing area at each data
collection location.
Location
16 17 18 19 20
Marked Intersection
Walk 23.71 9.70 — 15.07 —
+traffic flow change — — — — —
Don’t Walk — — — — —
+traffic flow change — — — — —
Overall in Intersection 20.99 8.70 — 13.58 —
Unmarked Non-Intersection
With Traffic 14.29 16.67 8.16 — —
+traffic flow change — — — — —
Against — 17.65 15.38 — —
+traffic flow change — — — — —
Overall in Unmarked Non-Intersection 11.11 17.07 9.60 — —
Grand Percentage 20.36 10.22 1.53 13.10 —
— Indicates no actions of this type were recorded.

159
Yielding Comparison

It is important to understand overall yielding behavior. However, because of the infrequency of

observed yielding in phase 3, it is not explored further here. However, a comparison of all
20 locations is examined later.

Evasive Pedestrian Actions

Table 42 summarizes the percentage of pedestrian evasive actions within each crossing area at
each of the data collection locations. The three types of evasive actions (running/accelerated
walking, abrupt stopping, and directional change) were combined to obtain a better overall
perspective on pedestrian evasive actions.

The percentages of crossings that involved pedestrian evasive actions were compared across
locations. The mean percentage of crossings that involved a pedestrian evasive action was 1.95.
None of the locations had a mean percentage that was more than 2 standard deviations away
from this mean. In other words, no outliers existed. These overall values do not provide much
information about where pedestrians are making evasive actions. As a result, evasive actions by
crossing area are examined next.

The mean percentage of pedestrians who took evasive actions in the marked intersection was
1.30. If all of the locations are compared with one another, none are considered outliers.

Next, the mean percentage of pedestrians who took evasive actions in unmarked non-
intersections was examined. Overall, the mean percentage was 4.39. No outliers exist here. In
other words, when looking at these five locations, none are significantly different with regard to
the proportion of pedestrians who took evasive actions in unmarked non-intersection areas.

A t-test was performed to determine whether a difference existed in the percentage of pedestrian
evasive actions between the marked intersection and the unmarked non-intersection. No
significant difference between the two locations was found, t(4) = -.818, p > .05. Interestingly,
none of the pedestrian evasive actions took place during the walk phase in the marked
intersection. Pedestrian evasive actions that took place in crossings that occurred within the
marked intersection entirely during the walk phase (M = 0.00 percent) were compared with all
other crossings (M = 7.28 percent). No significant difference was found, t(4) = -1.57, p > .05.

160
 Table 42. Percentage of pedestrian evasive actions in each crossing area at each data
collection location.
Location
16 17 18 19 20
Marked Intersection
Walk — — — — —
+traffic flow change 50.00 — — — —
Don’t Walk 37.50 7.69 — — 5.00
+traffic flow change 16.67 — — — —
Overall in Intersection 4.58 .54 — — 1.37
Unmarked Non-Intersection
With Traffic — — 3.06 — 33.33
+traffic flow change — — — — —
Against Traffic — — 3.85 — 18.18
+traffic flow change — — — — —
Overall in Unmarked Non-Intersection — — 3.20 — 18.75
Grand Percentage 4.29 .44 .51 — 4.49
— Indicates no actions of this type were recorded.

Evasive Vehicle Actions

Table 43 summarizes the percentage of vehicle evasive actions within each crossing area at each
of the data collection locations. Each of the four types of evasive actions (abrupt braking—first
vehicle, abrupt braking—second vehicle, directional change—first vehicle, and directional
change—second vehicle) were combined to obtain a better overall perspective on vehicle evasive
actions.

Location 17 was the only location where evasive actions were observed. As a result, no further
analyses were completed.

161
Table 43. Percentage of vehicle evasive actions in each crossing area at each data collection
location.
Location
16 17 18 19 20
Marked Intersection
Walk — .61 — — —
+traffic flow change — — — — —
Don’t Walk — — — — —
+traffic flow change — — — — —
Overall in Intersection — .54 — — -
Unmarked Non-Intersection
With Traffic — — — — —
+traffic flow change — — — — —
Against Traffic — — — - —
+traffic flow change — — — — —
Overall in Unmarked Non-Intersection — — — — —
Grand Percentage — .44 — — —
— Indicates no actions of this type were recorded.

Evasive Action Comparison

Because no vehicles took evasive actions, a comparison between vehicle and pedestrian evasive
actions was not made.

DISCUSSION

An overarching goal of the present project was to determine which environmental factors
influence where pedestrians cross the roadway. Pedestrian crossing behaviors were observed and
coded at five different locations in this phase. It was hoped that these data would help to identify
factors that influence pedestrians to cross at unmarked non-intersection locations. As was
discovered in phase 1 of this study, the number of unmarked non-intersection crossings is overall
quite low. As a result, modeling was not attempted for the phase 3 sites alone. Instead all data
were combined to create a predictive model. It is discussed in more detail later.

In phase 3, none of the locations were identified as outliers in terms of crossing location,
yielding, or evasive actions. Next, all 20 locations (phases 1–3) are compared with one another
and environmental factors are explored.

162
 PEDESTRIAN CROSSING BEHAVIORS: COMBINED RESULTS

In all three phases of this study, pedestrian crossing behaviors were examined. Although phase 1
and phase 2 used video capture and phase 3 used in-person observations, the same pedestrian
crossing coding guidelines were used in each phase. As a result, crossings are compared across
the 20 different locations here. In addition, for ease of reading, data from all three phases are
combined and presented again in this section.

The relationship between pedestrian crossings and various environmental factors was examined.
A summary of continuous variables can be found in table 44. A summary of categorical variables
can be found in table 45. The environmental variables are as follows:

 Distance (in ft) between marked crossings. This is the distance between the two
marked pedestrian crosswalks at each location.

 AADT. Average annual daily traffic expressed in thousands and rounded to the nearest
100.

 Trip originators. A scale from 1 (very few) to 5 (a lot) that quantifies the attractions in
the area that might generate pedestrian traffic.

 Walk phase. The length (in s) that the walk sign is illuminated (this includes the flashing
portion of the walk phase that warns pedestrians not to start crossing).

 Don’t walk phase. The length (in s) that the don’t walk is illuminated (steady state stable
sign only).

 Width of crossing. The distance (in ft) pedestrians are required to travel in the roadway
to cross in the marked intersection at each location.

 Travel pace. The rate (in ft/s) at which pedestrians are required to travel to complete a
crossing in the marked intersection entirely during the walk phase.

 Traffic direction. Indicator of one-way or two-way vehicular traffic 1 = one way and
2 = two way.

 Barriers. Indicator of barrier blocking all or part of the sidewalk from the roadway;
0 = none, 1 = partial block, and 2 = mostly blocked.

 Bus stops. Indicator of the location of bus stops; 0 = no bus stop, 1 = enter/exit bus at
crosswalk, and 2 = enter/exit bus at non-intersection.

 Parking present. Indicator whether parking was present on either side of the street (yes
or no).

 Center turn lane. Indicator whether a center (non-through traffic) lane was present (yes
or no).

163
  Dedicated right turn. Indicator whether a right turn (non-through traffic) lane was
present (yes or no).

 Median. Indicator of the type of median if it was present; 0 = no median, 1 = hard, raised
above traffic level median, and 2 = soft, not raised median (i.e., paint only).

 Cross streets. Indicator of a cross street that met with the main travel road between the
two marked intersections; 0 = no cross street, 1 = vehicle traffic was light controlled at
the cross street, 2 = vehicle traffic was not light controlled at the cross street.

 Far intersection light controlled. Indicator whether the far intersection (i.e., the next
marked crosswalk) was light controlled (yes or no).

Table 44. Summary of pedestrian crossing locations and their associated environmental
factors that were assessed on a continuous scale.
Distance
Between Width of
Marked Trip Walk Don’t Crossing Travel
Location Crossings (ft) AADT Originators Phase (s) Walk (s) (ft) Pace (ft/s)
1 352 15.8 1 20 69 61 3.1
2 530 24.2 1 46 55 81.25 1.8
3 550 26.3 1 46 71 48.25 1.0
4 656 11.3 2 32 45 54.42 1.7
5 391 15.1 3 53 50 61 1.2
6 294 31.5 3 31 56 83 2.7
7 145 11.9 1 44 52 43 1.0
8 433 24.2 1 19 79 70 3.7
9 551 34 4 30 68 109 3.6
10 361 41.8 3 24 75 72.5 3.0
11 193 22.9 1 19 80 68.3 3.6
12 277 19.8 3.5 20 78 64.5 3.2
13 316 15.1 1 10 89 50 5.0
14 338 34.7 1 30 67 68.25 2.3
15 297.5 18.3 1 39 60 45.5 1.2
16 361 28 4 58 73 75 1.3
17 266.5 8.1 5 64 63 43.25 0.7
18 511.25 7.9 4.5 60 40 30 0.5
19 342 23.2 2 30 70 80 2.7
20 339.5 13 3.5 20 63 44.5 2.2

164
 Table 45. Summary of pedestrian crossing locations and their associated environmental
factors that were assessed on a categorical scale.
Far
Center Dedicated Intersection
Traffic Bus Parking Turn Right Cross Light
Location Direction Barriers Stops Present Lane Turn Median Streets Controlled
1 2 2 1 Yes Yes No 0 0 Yes
2 2 0 0 Yes No Yes 2 0 Yes
3 2 1 2 No Yes No 1 1 Yes
4 1 0 2 Yes No No 0 0 Yes
5 1 0 0 Yes No Yes 0 0 Yes
6 2 0 0 No No No 0 0 Yes
7 1 1 1 Yes No No 0 0 No
8 2 0 0 No No No 2 2 No
9 2 0 2 Yes No Yes 1 2 Yes
10 2 0 0 Yes No No 0 0 Yes
11 2 0 1 No No No 0 0 No
12 2 0 1 Yes No No 0 0 Yes
13 1 0 2 Yes No No 0 0 Yes
14 2 0 0 No No No 0 0 Yes
15 1 0 2 No No No 0 0 Yes
16 2 0 1 Yes No No 1 0 Yes
17 2 0 2 Yes No Yes 0 0 Yes
18 1 0 1 No No No 0 0 No
19 2 2 2 No No No 1 0 Yes
20 1 0 0 Yes No Yes 0 0 Yes

The data were analyzed in two different ways. First, data were examined as the full raw dataset,
which includes more than 60,000 pedestrian crossings. Second, to handle different levels of
pedestrian traffic, data were analyzed in terms of the proportion of crossings within each
location.

165
Full Dataset

Overview of Crossings

Figure 61 through figure 66 depict the breakdown of the total observed crossings across all
20 locations. Because all of the crossings are combined across locations, the proportions of each
type of crossing may not be typical of all locations. Because Location 3 was the only location
that included crossings at an unmarked intersection, the sum values are not the same in every
figure.

2,322 Total Crossings

5,218

7,798 Marked Intersection - Rule-

Following Crossings
Marked Intersection - Rule-
Breaking Crossings
Unmarked Non-Intersection
Crossings
Unmarked Intersection Crossings

55,040

Figure 61. Chart. Distribution of all crossings observed across the 20 different locations by
the area in which they took place.

166
 Marked Intersection Crossings

Unimpeded crossing
Vehicle yielding
9,627
51,489 11,349 Pedestrian evasive action
Pedestrian yielding
Vehicle evasive action

1,547
163
12

Figure 62. Chart. Distribution of crossings observed across all 20 locations in the marked
intersections by circumstances of the crossing.

Unmarked Non-Intersection Crossings

Unimpeded crossing
341
Pedestrian evasive action
4,707 511 Pedestrian yielding
Vehicle yielding
Vehicle evasive action
97

70
3

Figure 63. Chart. Distribution of crossings observed across all 20 locations in the
unmarked non-intersections by the circumstances of the crossing.

167
 Rule-Breaking Crossings

Unimpeded crossing
1,063 Pedestrian evasive action
13,911 1,427 Pedestrian yielding
Vehicle yielding
Vehicle evasive action
205
155
4

Figure 64. Chart. Distribution of crossings observed across all 20 locations for all rule-
breaking crossings by the circumstances of the crossing (including those in the unmarked
non-intersections).

Marked Intersection - Rule-Following

Crossings

Unimpeded crossing
Vehicle yielding
9,559
44,448 10,592 Pedestrian evasive action
Pedestrian yielding
Vehicle evasive action

891
131
11

Figure 65. Chart. Distribution of crossings observed across all 20 locations for crossings
made entirely during the walk light phase in the marked intersections, by the
circumstances of the crossing.

168
 Marked Intersection - Rule-Breaking Crossings

Unimpeded crossing
Pedestrian evasive action
656
7,041 757 Vehicle yielding
Pedestrian yielding
Vehicle evasive action

68
32 1

Figure 66. Chart. Distribution of crossings observed across all 20 locations for rule-
breaking crossings in marked intersections (i.e., crossings that took place at least partially
during the don’t walk light phase) by the circumstances of the crossing.

Prediction Model—Crossing Area

Using data from all 20 locations, a model was successfully created in SAS® to predict whether
pedestrians would cross at the marked intersection (and not the unmarked non-intersections).
There were 68,056 pedestrian crossings among the 20 locations, 62,838 of which occurred at
marked intersections. (Note that crossings at the unmarked intersection at Location 3 were not
included in the model.) In other words, approximately 92 percent of the crossings occurred at the
marked intersections.

First, a factor analysis with an orthogonal varimax rotation was used to describe the underlying
relationships among the 16 environmental variables (see table 46). Based on the greater-than-one
rule for the eigenvalues, five factors were kept for the rotation. The factor loadings are included.
Combined, these factors accounted for approximately 74 percent of the standardized variance in
the data.

169
 Table 46. Environmental factors and their labels used to calculate the crossing location
prediction model.
Label Environmental Variable Coding
A Distance to the next marked crosswalk Distance in ft
B AADT Expressed in thousands and rounded to the
nearest 100
C One-way or two-way street 1 or 2
D Presence of physical barriers that might No barrier (0), partial barrier (1), or mostly
prevent a pedestrian from crossing the blocked/large barrier (2)
roadway
E Presence of a bus stop None (0), bus exit near marked intersection
(1), bus exit at Non-Intersection (2)
F Range of the number of trip originators/ Range from very few (1) to a lot (5)
destinations
G Presence of parking along the roadway Yes (1) or no (0)
H Presence of a center turning lane Yes (1) or no (0)
I Presence of a right turn only turning lane Yes (1) or no (0)
J Length of walk phase Time in s
K Length of don’t walk phase Time in s
L Curb-to-curb distance Distance in ft
M Presence and type of median No median (0), soft (1), hard (2), soft, not
raised median
N Presence of cross streets between marked No cross street (0), light-controlled cross
crosswalks. street (1), not light-controlled cross street
(2)
O Far marked crosswalk light controlled Yes (1) or no (0)
P Travel pace ft/s
Note: Values in parentheses are the values assigned to categorical variables.

170
 Table 47. Factor loadings for the 16 environmental variables.
Variable Factor 1 Factor 2 Factor 3 Factor 4 Factor 5
A -23 -4 74 13 21
B 24 86 11 -9 -6
C 11 86 13 19 -3
D 2 6 -9 82 -14
E 7 -39 17 51 16
F -40 4 -4 -36 51
G 10 -19 -7 -7 79
H -2 6 10 86 -5
I -25 -3 29 -30 65
J -92 -5 6 -2 9
K 78 33 -3 13 -2
L 28 75 34 -13 21
M -2 38 76 3 -13
N 23 16 84 -3 -12
O 2 29 -12 29 75
P 94 19 8 -8 2
 Indicates a loading of 60 or greater in absolute value.

Each of the five factors is assigned a descriptive label based on its correlations with relative
variables:

 Factor 1, Travel Pace and Phasing, is negatively correlated with the length of the walk
phase, positively correlated with the length of the do not walk phase, and positively
correlated with the travel pace.

 Factor 2, Traffic Throughput, is positively correlated with the AADT, traffic

directionality (one- or two-way street), and the curb-to-curb distance (i.e., the width of
the street).

 Factor 3, Distance to Safety, is positively correlated with the distance to the next marked
crosswalk, the presence and type of median, and the presence of cross streets between
marked crosswalks.

 Factor 4, External Objects (Barriers/Vehicles) in Center of Road, is positively correlated

with the presence of physical barriers that might prevent a pedestrian from crossing the
roadway and the presence of a center turning lane.

 Factor 5, External Objects (Vehicles) on Sides of Road, is positively correlated with the
presence of parking along the roadway, the presence of a right turn only lane, and
whether or not the far marked crosswalk is light controlled.

Scores for each of the five factors were generated using the standardized scoring coefficients
listed in table 48. These coefficients can only be used with the standardized data and not with the

171
raw, observed values. Each of the five factor scores is graphed against the percentage of
intersection crossings for each location in figure 67.

Table 48. Standardized scoring coefficients for the five rotated factors.
Variable Factor 1 Factor 2 Factor 3 Factor 4 Factor 5
A -0.05047 -0.10759 0.38127 0.07187 0.08558
B -0.04080 0.36229 -0.07173 -0.04503 -0.03821
C -0.09979 0.38416 -0.07152 0.10124 -0.01225
D -0.04177 0.05360 -0.06575 0.39620 -0.01480
E 0.09552 -0.22544 0.14359 0.25108 0.12869
F -0.12612 0.07613 -0.04661 -0.12670 0.20952
G 0.12938 -0.11604 -0.01154 0.01230 0.41212
H -0.04871 0.03214 0.02615 0.41759 0.02701
I -0.03224 -0.03729 0.13973 -0.10183 0.29012
J -0.36946 0.11821 0.00329 0.02802 -0.01769
K 0.26616 0.03607 -0.04015 0.04254 0.04615
L 0.02736 0.25549 0.06317 -0.05440 0.09691
M -0.05098 0.06517 0.33786 -0.00163 -0.08623
N 0.09721 -0.09803 0.42527 -0.04528 -0.06540
O -0.00290 0.14773 -0.12528 0.19947 0.39702
P 0.37272 -0.08072 0.04904 -0.06944 0.06522

172
Notes: The dots remain in consistent locations from box to box.
solid line = regression line, shaded area = 95-percent confidence limits of the regression line).
Figure 67. Chart. Scatterplot of each location’s score for each of the five factors against the
percentage of intersection crossings at that location.

Next, the scores for the five factors were regressed on the probability of an intersection crossing
through logistic regression using forward selection. The model iterated three times after the
2 2
entrance of an intercept term. The scores for Factor 1, χ (1) = 176.6, p < 0.01, Factor 2, χ (1) =
2
41.6, p < 0.01, and Factor 3, χ (1) = 402.3, p < 0.01, were statistically significant. The resultant
model is shown in figure 68 where π represents the probability of crossing in the marked
intersection and x is the vector of factor scores for a given location. When using a logit (log
odds) model, eα represents the odds of success when all predictors equal zero, and eβ represents
the multiplicative effect of a one-unit increase in the corresponding x on the odds of success,
when the other predictors are held fixed.

Figure 68. Equation. The logit of crossing at the intersection.

A good probability cutoff point for this data is 0.9, meaning that any future predictions using the
model that are 0.9 or greater should be deemed an intersection crossing. With this cutoff, the

173
model can successfully predict approximately 90 percent of the crossings overall. The resultant
sensitivity and specificity are about 96 percent and 10 percent, respectively. Sensitivity measures
how many events (intersection crossings) were successfully predicted, and specificity measures
how many non-events (non-intersection crossings) were successfully predicted. The predicted
probabilities of crossing at the intersection for each location are shown in table 49.

Table 49. Probabilities of correctly predicting a crossing at the marked intersection or the
unmarked non-intersection by location and the corresponding upper and lower 95-percent
confidence limits.
Lower Upper
95 Percent 95 Percent
Total Estimated Confidence Confidence
Location Crossings Probability Limit Limit
1 1,110 0.91203 0.90715 0.91668
2 4,631 0.90319 0.89779 0.90834
3 2,878 0.90348 0.89874 0.90802
4 13,199 0.91365 0.90995 0.91721
5 10,635 0.94112 0.93862 0.94353
6 16,418 0.93063 0.92710 0.93400
7 12,958 0.94419 0.94152 0.94675
8 1,574 0.80553 0.79210 0.81829
9 805 0.80876 0.79562 0.82123
10 528 0.91159 0.90585 0.91701
11 17 0.89962 0.89213 0.90664
12 185 0.90793 0.90141 0.91406
13 609 0.85845 0.84056 0.87463
14 205 0.92484 0.92110 0.92842
15 840 0.93348 0.93106 0.93583
16 280 0.93036 0.92719 0.93340
17 225 0.94899 0.94636 0.95150
18 786 0.95218 0.94816 0.95590
19 84 0.90816 0.90462 0.91158
20 89 0.92066 0.91693 0.92424

Prediction Model—Walk Versus All Other Crossings

A second model was developed to predict crossings that occurred entirely during the walk phase
in the marked intersection. Crossings were divided into two categories—crossings that occurred
within the marked intersection entirely during the walk light phase and all other crossings (e.g.,
unmarked non-intersection, marked intersection during the don’t walk light phase, etc.). Using
this classification system, there were 70,378 pedestrian crossings among the 20 locations, 55,040
of which occurred entirely within the walk phase in the marked intersection. In other words,
approximately 78 percent of the crossings were rule-following crossings. Using the factor
analysis and standardized scoring coefficients (table 48), each of the five factor scores is graphed
against the percentage of rule-following crossings for each location in figure 69.

174
Notes: The dots remain in consistent locations from box to box.
solid line = regression line, shaded area = 95-percent confidence limits of the regression line).
Figure 69. Chart. Scatterplot of each location’s score for each of the five factors against the
percentage of intersection crossings at that location.

Next, the scores for the five factors were regressed on the probability of a rule-following
crossing through logistic regression using forward selection (see figure 70). The model iterated
2
five times after the entrance of an intercept term. The scores for Factor 1, χ (1) = 607.6, p < 0.01,
2 2 2
Factor 2, χ (1) = 1224.5, p < 0.01, Factor 3, χ (1) = 187.4, Factor 4, χ (1) = 5783.2, p < 0.01,
2
and Factor 5, χ (1) = 388.4, p < 0.01, were all statistically significant.

Figure 70. Equation. The logit of crossing at the intersection during the walk phase.

A good probability cutoff point for this data is 0.6, meaning that any future predictions using the
model that are 0.6 or greater should be deemed a rule-following crossing. With this cutoff, the
model can successfully predict approximately 79 percent of the crossings overall. The resultant

175
sensitivity and specificity are about 94 percent and 23 percent, respectively. The predicted
probabilities of crossing during the walk phase at the intersection for each location are shown in
table 50.

Table 50. Probabilities of correctly predicting a crossing at the marked intersection entirely
during the walk phase by location and the corresponding upper and lower 95-percent
confidence limits.
Lower Upper
95 Percent 95 Percent
Total Estimated Confidence Confidence
Location Crossings Probability Limit Limit
1 1,110 0.7514 0.7251 0.7759
2 4,631 0.8791 0.8694 0.8882
3 2,878 0.4456 0.4321 0.4591
4 13,199 0.6798 0.6718 0.6877
5 10,635 0.9343 0.9295 0.9389
6 16,418 0.9062 0.9017 0.9106
7 12,958 0.7546 0.7471 0.7619
8 1,574 0.7421 0.7199 0.7631
9 805 0.7764 0.7463 0.8039
10 528 0.8431 0.8106 0.8710
11 17 0.6428 0.5546 0.7223
12 185 0.6431 0.5752 0.7057
13 609 0.2855 0.2530 0.3204
14 205 0.8054 0.7656 0.8399
15 840 0.5515 0.5181 0.5845
16 280 0.8286 0.7798 0.8683
17 225 0.7344 0.6730 0.7879
18 786 0.7304 0.6983 0.7603
19 84 0.8690 0.7788 0.9260
20 89 0.4227 0.3397 0.5103

Crossing Proportions

To handle different levels of pedestrian traffic, the next set of analyses examine the data in terms
of the proportion of crossings at each location in this section. The proportion of crossings and the
types of crossings at each location are examined as a complete group here. In addition, the
relationship between environmental factors and the proportions of crossing types are examined.

Crossing Location

Table 51 summarizes the percentage of pedestrians, by location, who crossed at a marked

intersection and unmarked non-intersection. An additional column, “rule-breaking,” lists the
percentage of people within each area whose crossing was not completed entirely during the
walk phase in the marked intersection (e.g., an unmarked non-intersection or a marked
intersection in the don’t walk phase). The percentage of each type of crossing is presented at

176
each location. For example, at Location 19, 86.90 percent of the total crossings took place in the
marked intersection during the walk phase. The mean percentage of pedestrians crossing in the
marked intersection is 83.88. Location 3 is considered an outlier, with a mean percentage of
50.88 pedestrians crossing at the marked intersection (2.70 standard deviations below the mean).
As was noted in phase 1, Location 3 is unique in that there is an unmarked intersection present
between the two marked crosswalks.

A mean percentage of 13.89 pedestrians crossed in an unmarked non-intersection area.

Location 15 is considered an outlier, with a mean percentage of 36.55 pedestrians crossing at the
unmarked non-intersection (2.34 standard deviations above the mean). Location 15 (as described
in phase 2) includes a large median that divides traffic. This median provides pedestrians with an
opportunity to travel on a more direct (and potentially faster) path between public transit and the
neighboring residential area. It is likely that this played a role in the proportion of unmarked non-
intersection crossings.

Next, to examine rule-following, crossings made entirely within the walk phase in the marked
intersection were compared with all other crossings. (These other crossings are referred to as
“rule-breaking.” However, note that these crossings are not necessarily against the rules and/or
laws in all states. Rather, it is a simple convention for ease of description.) A mean percentage of
70.89 pedestrians crossed entirely during the walk phase in the marked intersections. Location 13
was an outlier at 28.41 percent (2.46 standard deviations below the mean).

A t-test was performed to confirm a difference in the percentage of pedestrians crossing the
marked intersections (M = 83.88 percent) and the unmarked non-intersections
(M = 13.89 percent). A significant difference between the two locations was found,
t(19)= 15.95, p < .001. A second t-test examined the percentage of crossings that occurred
entirely during the walk phase (M = 70.89 percent) and the percentage that occurred during rule
breaking (M = 29.11 percent) crossings. Indeed a difference between the groups was found,
t(19) = 5.40, p < .001.

177
 Table 51. Percentage of pedestrians crossing at each area in each data collection location.
Marked Intersection Unmarked Non-Intersection
Walk Don’t Walk With Traffic Against Traffic
Overall in
without without without without Rule-
Location +traffic +traffic Overall in +traffic +traffic Unmarked
traffic traffic traffic traffic Breaking
flow flow Intersection flow flow Non-
flow flow flow flow
change change change change Intersection
change change change change
1 75.14 0.72 8.47 3.60 87.93 5.23 0.45 5.77 0.63 12.07 24.86
2 87.91 1.12 1.75 0.52 91.30 6.24 0.30 1.60 0.56 8.70 12.09
3 44.56 0.38 5.73 0.21 50.88 1.00 0.15 3.15 0.15 4.46 55.44
4 67.98 0.71 17.89 4.59 91.17 5.38 0.17 2.77 0.52 8.83 32.02
5 93.43 0.37 0.49 0.05 94.33 5.19 0.14 0.33 0.01 5.67 6.57
6 90.63 0.82 0.92 0.90 93.26 4.67 0.27 1.11 0.68 6.74 9.37
7 75.46 1.30 18.69 1.53 96.98 1.81 0.01 1.20 0.00 3.02 24.54
8 74.21 0.64 1.33 0.64 76.81 16.58 0.64 5.46 0.51 23.19 25.79
9 77.64 2.48 4.10 4.84 89.07 3.60 0.50 6.58 0.25 10.93 22.36
10 84.28 0.00 0.00 0.19 84.47 8.71 0.19 6.44 0.19 15.53 15.72
11 58.82 0.00 11.76 0.00 70.59 17.65 0.00 11.76 0.00 29.41 41.18
12 64.86 0.54 3.24 0.00 68.65 15.68 1.08 14.59 0.00 31.35 35.14
13 28.41 10.84 46.80 4.93 90.97 5.09 0.33 3.61 0.00 9.03 71.59
14 80.49 1.46 4.88 2.93 89.76 6.34 0.00 3.41 0.49 10.24 19.51
15 55.24 0.36 6.79 1.07 63.45 28.21 0.83 7.38 0.12 36.55 44.76
16 82.86 5.71 2.86 2.14 93.57 5.00 0.00 0.36 1.07 6.43 17.14
17 73.33 1.78 5.78 0.89 81.78 10.67 0.00 7.56 0.00 18.22 26.67
18 73.03 5.47 4.20 1.40 84.10 12.47 0.13 3.31 0.00 15.90 26.97
19 86.90 8.33 0.00 1.19 96.43 1.19 0.00 1.19 1.19 3.57 13.10
20 42.70 2.25 22.47 14.61 82.02 3.37 0.00 12.36 2.25 17.98 57.30

178
To better understand influences on crossing behaviors, the relationship between crossing area
and the aforementioned environmental factors were examined. Only significant relationships are
discussed. A significant correlation was found between the width of the crossing and the
percentage of pedestrians crossing entirely during the walk phase in the marked intersection,
r(18) = .504, p = .024. In other words, the greater the travel distance, the more likely pedestrians
were to be rule followers; or the shorter the travel distance, the more likely pedestrians were to
be “rule breakers.”

Interestingly, there was also a significant negative relationship between the length of the walk
phase and crossing in the unmarked non-intersection against traffic, r(18) = -.456, p = .043. In
other words, the shorter the time available to cross during the walk phase, the more likely
pedestrians were to cross in the unmarked non-intersection against traffic flow.

The presence of physical barriers that might block all or part of the sidewalk from the roadway
(away from the marked intersection) were significantly negatively correlated with crossing in the
unmarked non-intersection, rs(18) = -.496, p = .026. In other words, with more physical barriers
present, pedestrians were less likely to cross in unmarked non-intersection areas.

Finally, traffic direction was significantly negatively correlated with crossings in the marked
intersection entirely during the don’t walk phase, rpb(18) = -.565, p = .005. In other words,
pedestrians were significantly more likely to cross entirely during the don’t walk phase on one-
way streets (16.76 percent) than on two-way streets (3.91 percent).

Pedestrian Yielding

Table 52 summarizes the percentage of pedestrians who yielded to vehicles within each crossing
type. The mean percentage of pedestrians yielding to vehicles across each of the 20 locations is
0.98. Location 17 was an outlier with a mean of 6.67 percent of pedestrian crossings involving
yielding to a vehicle (3.60 standard deviations above the mean). This is the result of a high
percentage of pedestrians yielding to turning vehicles while crossing in the marked intersection.
This is discussed in more detail later.

The mean percentage of pedestrians who yielded to vehicles in the marked intersection was 0.41.
Not surprisingly, Location 17 was an outlier again with a percentage of 6.52 (4.22 standard
deviations above the mean). All of these yielding behaviors were observed while pedestrians
crossed entirely during the walk phase.

Next, the mean percentages of pedestrians who yielded to vehicles in unmarked non-
intersections were examined. Overall, the mean percentage of yielding was 3.77. Location 20 is
considered an outlier at 12.50 percent (2.03 standard deviations above the mean). However,
given that there were only two instances of pedestrians yielding in the unmarked non-
intersection, this is not discussed further.

Next, to examine rule-following, crossings made entirely within the walk phase in the marked
intersection were compared with all other crossings. A mean percentage of 0.43 pedestrians
yielded while crossing entirely during the walk phase in the marked intersections. As previously
mentioned, Location 17 was an outlier at 7.27 percent (4.22 standard deviations above the mean).
Overall, a mean percentage of 2.58 pedestrians yielded to vehicles while performing a rule-

179
breaking crossing. Location 10 was an outlier at 13.25 percent (3.12 standard deviations above
the mean).

A simple t-test was performed to determine whether the proportion of pedestrian yielding
behaviors differed between the marked intersection (M = 0.41) and the unmarked non-
intersection (M = 3.77). A significant difference between the groups was found, t(19) = -3.52,
p = .002. A second t-test examined the percentage of yielding behaviors that occurred entirely
during the walk phase (M = 0.43) and the percentage that occurred during “rule breaking”
(M = 2.58) crossings. Indeed a difference between the groups was found, t(19) = -2.69,
p = 0.014.

180
 Table 52. Percentage of pedestrians yielding to vehicles in each crossing area at each data collection location.
Marked Intersection Unmarked Non-Intersection
Walk Don’t Walk With Traffic Against Traffic
Overall in
Without Without Without Without Rule- Grand
Location +traffic +traffic Overall in +traffic +traffic Unmarked
traffic traffic traffic traffic Breaking Percentage
flow flow Intersection flow flow Non-
flow flow flow flow
change change change change Intersection
change change change change
1 0.12 0.00 0.00 0.00 0.10 3.45 20.00 0.00 14.29 2.99 1.45 0.45
2 0.12 1.92 0.00 25.00 0.28 2.08 7.14 12.16 34.62 6.20 5.71 0.80
3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 5.49 25.00 4.74 3.02 1.67
4 0.17 2.13 0.00 0.50 0.17 0.14 0.00 0.00 0.00 0.09 0.14 0.16
5 0.03 0.00 0.00 0.00 0.03 0.18 0.00 2.86 0.00 0.33 0.29 0.05
6 0.15 0.00 0.00 4.76 0.19 0.65 2.22 8.20 4.50 2.35 2.14 0.33
7 0.74 4.14 0.17 0.51 0.67 0.43 0.00 0.65 0.00 0.51 0.44 0.66
8 0.09 0.00 0.00 0.00 0.08 0.00 10.00 0.00 0.00 0.27 0.25 0.13
9 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3.77 0.00 2.27 1.11 0.25
10 0.00 0.00 0.00 100.00 0.22 2.17 0.00 26.47 0.00 12.20 13.25 2.08
11 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
12 0.00 0.00 0.00 0.00 0.00 0.00 0.00 18.52 0.00 8.62 7.69 2.70
13 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
14 0.00 0.00 0.00 0.00 0.00 0.00 0.00 28.57 0.00 9.52 5.00 0.98
15 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
16 0.00 0.00 0.00 0.00 0.00 7.14 0.00 0.00 0.00 5.56 2.08 0.36
17 7.27 0.00 0.00 0.00 6.52 4.17 0.00 11.76 0.00 7.32 5.00 6.67
18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20 0.00 0.00 0.00 0.00 0.00 33.33 0.00 9.09 0.00 12.50 3.92 2.25

181
To better understand influences on crossing behaviors, the relationship between crossing
location, pedestrian yielding, and the aforementioned environmental factors were examined. A
significant correlation was found between the length of the walk phase and the percentage of
pedestrians yielding to vehicles both in the marked intersection, r(18) = .450, p = .047, and
crossings that took place entirely during the walk phase in the marked intersection, r(18) = .450,
p = .047. Vehicle traffic, as estimated by AADT, was also significantly correlated with
pedestrians yielding during rule-breaking crossings, r(18) = .478, p = .033.

A significant negative correlation was found between the length of the don’t walk phase and
pedestrian yielding when the crossing in the marked intersection began during the walk phase,
but ended in the don’t walk phase, r(18) = -.451, p = .046.

Several significant correlations were also found between categorical environmental variables and
pedestrian yielding locations. There was a significant correlation between whether the next
intersection was light controlled (M = 0.00 percent) or stop sign controlled (M = 0.04 percent)
and pedestrian yielding to vehicles while crossing entirely during the don’t walk phase in the
marked intersection, rpb(18) = -.459, p = .042.

Bus stop location was also correlated with pedestrians yielding while crossing in the unmarked
non-intersection against traffic, rpb(18) = -.452, p = .045. The mean percentages of pedestrian
yielding in this scenario were as follows: no bus stop is 12.48 percent, bus stop near the
crosswalk is 3.19 percent, and bus stop away from the intersection is 3.00 percent.

Vehicle Yielding

Table 53 summarizes the percentage of vehicles that yielded to pedestrians within each crossing
area. The overall percentages of crossings that involved a vehicle yielding to a pedestrian were
compared across locations. The mean percentage of crossings that involved vehicle yielding was
8.93. Location 7 was an outlier with 38.69 percent of the crossings involving a vehicle yielding
to a pedestrian (2.77 standard deviations above the mean). As was discussed in phase 1, there is
both high pedestrian traffic and turning vehicle traffic at this intersection. It is this combination
that likely led to the high percentage of vehicles yielding to pedestrians.

The mean percentage of vehicles that yielded to pedestrians in the marked intersection was 9.25.
Once again, Location 7 was considered to be an outlier, with 39.80 percent of the crossings in the
marked intersection including vehicles yielding to pedestrians (2.64 standard deviations above
the mean).

Next, the mean percentage of drivers who yielded to pedestrians in unmarked non-intersections
was examined. Overall, the mean percentage of yielding was 3.60. Location 17 was considered
an outlier with 17.07 percent (2.80 standard deviations above the mean).

Next, rule-following was examined. A mean of 11.00 percent yielded to pedestrians crossing
entirely during the walk phase in the marked intersections. As previously mentioned, Location 7
was an outlier at 50.71 percent (2.87 standard deviations above the mean). Overall, a mean
percentage of 2.55 vehicles yielded to pedestrians performing a rule-breaking crossing. Both
Location 17 at 11.67 percent (2.77 standard deviations above the mean) and Location 12 at
9.23 percent (2.03 standard deviations above the mean) were classified as outliers.

182
A t-test was performed to determine whether a difference existed in the percentage of vehicle
yielding behaviors between the marked intersections (M = 9.25 percent) and the unmarked non-
intersections (M = 3.60 percent). A significant difference between the two locations was found,
t(19) = 2.16, p =.044. A second t-test examined the percentage of yielding behaviors that
occurred entirely during the walk phase (M = 11.00 percent) and the percentage that occurred
during rule-breaking (M = 2.55 percent) crossings. Indeed, a difference between the groups was
found, t(19) = 2.72, p < .014.

183
 Table 53. Percentage of vehicles yielding to pedestrians in each crossing area at each data collection location.
Marked Intersection Unmarked Non-Intersection
Walk Don’t Walk With Traffic Against Traffic
Overall in
Without Without Without Without Rule- Grand
Location +traffic +traffic Overall in +traffic +traffic Unmarked
traffic traffic traffic traffic Breaking Percentage
flow flow Intersection flow flow Non-
flow flow flow flow
change change change change Intersection
change change change change
1 4.08 0.00 0.00 0.00 3.48 5.17 0.00 4.69 0.00 4.48 2.17 3.60
2 7.22 0.00 0.00 0.00 6.95 0.00 0.00 0.00 0.00 0.00 0.00 6.35
3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.61 0.00 0.43 0.10 0.35
4 29.88 0.00 0.00 0.99 22.33 0.70 0.00 0.27 0.00 0.52 0.28 20.40
5 0.01 28.21 0.00 0.00 0.12 0.18 6.67 0.00 0.00 0.33 1.86 0.13
6 8.74 0.74 0.00 0.00 8.50 0.13 0.00 0.00 0.00 0.09 0.13 7.93
7 50.71 16.57 0.62 0.51 39.80 2.13 0.00 4.52 0.00 3.07 1.76 38.69
8 0.34 20.00 0.00 0.00 0.50 0.38 10.00 2.33 0.00 1.10 1.48 0.64
9 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.89 0.00 1.14 0.56 0.12
10 32.81 0.00 0.00 0.00 32.74 0.00 0.00 11.76 0.00 4.88 4.82 28.41
11 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
12 10.00 0.00 0.00 0.00 9.45 0.00 0.00 22.22 0.00 10.34 9.23 9.73
13 13.87 0.00 0.00 0.00 4.33 0.00 0.00 4.55 0.00 1.82 0.23 4.11
14 13.94 33.33 0.00 0.00 13.04 7.69 0.00 0.00 0.00 4.76 5.00 12.20
15 0.00 100.00 0.00 0.00 0.56 0.00 0.00 6.45 0.00 1.30 1.86 0.83
16 23.71 0.00 0.00 0.00 20.99 14.29 0.00 0.00 0.00 11.11 4.17 20.36
17 9.70 0.00 0.00 0.00 8.70 16.67 0.00 17.65 0.00 17.07 11.67 10.22
18 0.00 0.00 0.00 0.00 0.00 8.16 0.00 15.38 0.00 9.60 5.66 1.53
19 15.07 0.00 0.00 0.00 13.58 0.00 0.00 0.00 0.00 0.00 0.00 13.10
20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

184
To better understand influences on crossing behaviors, the relationship among crossing location,
vehicle yielding, and the aforementioned environmental factors were examined. There were
several significant relationships between pedestrian crossing location and environmental factors
that included vehicle yielding. First, a significant relationship between the number of trip
originators and destinations and pedestrians completing rule-breaking crossings, r(18) = .586,
p = .009 was found. In other words, when looking at instances where pedestrians crossed in a
manner other than entirely during the walk phase in the marked intersection, there were more
instances of vehicles yielding to pedestrians as the number of trip originators increased.

The length of the walk phase was significantly correlated with the percentage of vehicles
yielding to pedestrians crossing in unmarked non-intersections, r(18) = .475, p = .034. A
significant negative relationship was also found between the length of the don’t walk light phase
and the percentage of vehicles yielding to pedestrians making crossings in the marked
intersection that began during the don’t walk light phase, but completed during the walk light
phase, r(18) = -.454., p = .044.

Interestingly, a significant negative correlation was found between the required pedestrian travel
pace and yielding to pedestrians crossing in the unmarked non-intersection area with traffic,
r(18), = -.456, p = .043. This suggests that the faster pedestrians were required to travel (i.e., less
time), the more likely they were to experience a vehicle yielding to them while crossing.

Only a single correlation with a categorical environmental variable was found here. A significant
relationship was found between vehicles yielding to pedestrians crossing entirely during the
don’t walk phase and whether the next intersection was light controlled (M = 0.00 percent) or
stop sign controlled (M = 0.15 percent), rpb(18) = -459, p = .042.

Yielding Comparison

It is important to understand overall yielding behavior. Here pedestrian and vehicle yielding are
compared. First, yielding behaviors within the marked intersection were compared. A
significantly greater percentage of crossings involved vehicle yielding (M = 9.25 percent) than
pedestrian yielding (M = 0.41 percent), t(19) = -3.41, p = .003. Next, yielding behaviors in
unmarked non-intersections were examined. No significant difference between vehicle
(M = 3.60 percent) and pedestrian (M = 3.77 percent) yielding was found, t(19) = .145, p > .05.

Next, rule following was examined. When pedestrians crossed in the marked intersection entirely
during the walk phase, significantly more vehicles yielded to pedestrians (M = 11.00 percent)
than pedestrians yielded to vehicles (M = .43 percent), t(19) = -3.41, p = .003. When pedestrians
made a rule-breaking crossing, there was no significant difference in the percentage of vehicles
that yielded to pedestrians (M = 2.57 percent) and pedestrians that yielded to vehicles
(M = 2.55 percent), t(19) = 0.03, p > .05.

Evasive Pedestrian Actions

Table 54 summarizes the percentage of pedestrian evasive actions within each crossing area at
each of the data collection locations. The three types of evasive actions (running/accelerated
walking, abrupt stopping, and directional change) were combined to obtain a better overall
perspective on pedestrian evasive actions.

185
The overall percentage of crossings that involved a pedestrian evasive action was compared
across locations. The mean percentage of crossings that involved a pedestrian evasive action was
3.06. Both Location 1 (8.92 percent; 2.23 standard deviations above the mean) and Location 8
(8.89 percent; 2.22 standard deviations above the mean) were considered to be outliers.

The mean percentage of pedestrians who took evasive actions in the marked intersection was
2.16. Once again, Location 1 was considered an outlier at 6.67 percent (2.29 standard deviations
above the mean). Next, the mean percentage of pedestrians who took evasive actions in
unmarked non-intersections was examined. Overall, the mean percentage was 7.47. Once again,
Location 1 was considered an outlier at 24.63 percent (2.40 standard deviations above the mean).

Next, rule-following was examined. A mean of 1.02 percent of pedestrians took evasive actions
while crossing entirely during the walk phase in the marked intersections. Location 2 was an
outlier at 4.96 percent (2.66 standard deviations above the mean). Overall, a mean percentage of
8.91 pedestrians took evasive actions while making a rule-breaking crossing. Location 16 was an
outlier at 25.00 percent (2.01 standard deviations above the mean).

A t-test was performed to determine whether a difference existed in the percentage of pedestrian
evasive actions between the marked intersections (M = 2.16 percent) and the unmarked non-
intersections (M = 7.47 percent). A significant difference between the two locations was found,
t(19) = -3.82 , p =.001. A second t-test examined the percentage of evasive actions that occurred
entirely during the walk phase (M = 1.02 percent) and the percentage that occurred during rule-
breaking (M = 8.91 percent) crossings. Indeed a difference between the groups was found,
t(19) = -4.50, p < .001.

186
 Table 54. Percentage of pedestrian evasive actions in each crossing area at each data collection location.
Marked Intersection Unmarked Non-Intersection
Walk Don’t Walk With Traffic Against Traffic
Overall in
Without Without Without Without Rule- Grand
Location +traffic +traffic Overall in +traffic +traffic Unmarked
traffic traffic traffic traffic Breaking Percentage
flow flow Intersection flow flow Non-
flow flow flow flow
change change change change Intersection
change change change change
1 3.72 12.50 28.72 17.50 6.76 18.97 0.00 31.25 28.57 24.63 24.64 8.92
2 4.96 25.00 19.75 16.67 5.56 4.84 0.00 21.62 26.92 9.18 12.50 5.87
3 0.22 20.00 7.05 0.00 1.13 3.85 0.00 9.15 12.50 7.76 0.94 2.19
4 1.59 12.77 4.57 1.49 2.26 1.41 0.00 1.92 0.00 1.46 3.45 2.19
5 1.70 74.36 36.54 40.00 2.18 2.17 20.00 14.29 0.00 3.32 10.01 2.25
6 1.83 28.15 25.83 7.48 2.36 3.00 24.44 11.48 5.41 5.52 9.68 2.57
7 0.24 14.79 8.42 0.51 2.01 4.68 0.00 14.84 0.00 8.70 8.30 2.21
8 3.42 80.00 42.86 70.00 5.29 13.79 40.00 40.70 12.50 20.82 24.63 8.89
9 0.00 0.00 9.09 2.56 0.56 0.00 0.00 5.66 0.00 3.41 3.89 0.87
10 0.45 0.00 0.00 0.00 0.45 2.17 0.00 23.53 0.00 10.98 10.84 2.08
11 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
12 1.67 0.00 33.33 0.00 3.15 0.00 50.00 22.22 0.00 12.07 13.85 5.95
13 0.58 3.03 4.56 3.33 3.07 0.00 0.00 13.64 0.00 5.45 4.36 3.28
14 0.00 0.00 20.00 0.00 1.09 15.38 0.00 0.00 0.00 9.52 10.00 1.95
15 0.00 0.00 5.26 11.11 0.75 0.84 0.00 19.35 0.00 4.56 4.79 2.14
16 0.00 50.00 37.50 16.67 4.58 0.00 0.00 0.00 0.00 0.00 25.00 4.29
17 0.00 0.00 7.69 0.00 0.54 0.00 0.00 0.00 0.00 0.00 1.67 0.44
18 0.00 0.00 0.00 0.00 0.00 3.06 0.00 3.85 0.00 3.20 1.89 0.51
19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20 0.00 0.00 5.00 0.00 1.37 33.33 0.00 18.18 0.00 18.75 7.84 4.49

187
To better understand influences on crossing behaviors, the relationship among crossing location,
pedestrian evasive actions, and the aforementioned environmental factors were examined. A
significant negative relationship between the length of the walk light phase and evasive actions
made by pedestrians crossing in unmarked non-intersections was found, r(18) = -4.94, p = .027.
In other words, the shorter the amount of time that pedestrians had to cross the roadway during
the walk phase, the more likely they were to make an evasive maneuver while crossing outside
the marked intersection.

A relationship between bus stop location and the percentage of pedestrians taking evasive actions
while making a rule-breaking crossing was found, rpb(18) = -.509, p =.022. The mean
percentages of pedestrian evasive actions in this scenario were: no bus stop = 12.22 percent, bus
stop near the crosswalk = 12.23 percent, and bus stop away from the intersection = 2.73 percent.

Evasive Vehicle Actions

Table 55 summarizes the percentage of vehicle evasive actions within each crossing area at each
of the data collection locations. Each of the four types of evasive actions (abrupt braking—first
vehicle, abrupt braking—second vehicle, directional change—first vehicle, and directional
change—second vehicle) were combined to obtain a better overall perspective on vehicle evasive
actions.

The overall percentage of crossings that involved a vehicle evasive action was compared across
locations. The mean percentage of crossings that involved a vehicle evasive action was 0.10.
Location 12 had 1.08 percent of the crossings involve a vehicle evasive action, which was an
outlier (3.82 standard deviations above the mean).

The mean percentage of pedestrians who took evasive actions in the marked intersection was
0.04. Location 17 was considered an outlier at 0.54 percent (3.83 standard deviations above the
mean). Next, the mean percentage of vehicles that took evasive actions in unmarked non-
intersections was examined. Overall, the mean percentage was 0.21. Once again, Location 12
was considered an outlier at 3.45 percent (4.15 standard deviations above the mean).

Next, rule-following was examined. A mean of 0.04 percent of vehicles took evasive actions
while pedestrians were crossing entirely during the walk phase in the marked intersections.
Location 17 was an outlier at .61 percent (3.93 standard deviations above the mean). Overall, a
mean percentage of 0.19 vehicles took evasive actions while a pedestrian completed a rule-
breaking crossing. Location 12 was an outlier at 3.08 percent (4.13 standard deviations above the
mean).

A t-test was performed to determine whether a difference existed in the percentage of vehicle
evasive actions between the marked intersections (M = 0.04 percent) and the unmarked non-
intersections (M = 0.21 percent). No significant difference between the two locations was found,
t(19) = -.929, p > .05. A second t-test examined the percentage of evasive actions that occurred
entirely during the walk phase (M = 0.04 percent) and the percentage that occurred during “rule
breaking” (M = 0.19 percent) crossings. No significant difference between the two locations was
found, t(19) = -.906, p > .05.

188
 Table 55. Percentage of vehicle evasive actions in each crossing area at each data collection location.
Marked Intersection Unmarked Non-Intersection
Walk Don’t Walk With Traffic Against Traffic
Overall in
Without Without Without Without Rule- Grand
Location +traffic +traffic Overall in +traffic +traffic Unmarked
traffic traffic traffic traffic Breaking Percentage
flow flow Intersection flow flow Non-
flow flow flow flow
change change change change Intersection
change change change change
1 0.00 0.00 1.06 0.00 0.10 1.72 0.00 0.00 0.00 0.75 0.72 0.18
2 0.25 0.00 0.00 0.00 0.24 0.00 0.00 0.00 0.00 0.00 0.00 0.22
3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
7 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
8 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
9 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
10 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
11 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
12 0.00 0.00 0.00 0.00 0.00 0.00 0.00 7.41 0.00 3.45 3.08 1.08
13 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
14 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
15 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
16 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
17 0.61 0.00 0.00 0.00 0.54 0.00 0.00 0.00 0.00 0.00 0.00 0.44
18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

189
To better understand influences on crossing behaviors, the relationship among crossing location,
vehicle evasive actions, and the aforementioned environmental factors were examined. A
significant correlation was found between the length of the walk phase and vehicle evasive
actions taken while pedestrians crossed entirely during the walk phase in the marked intersection,
r(18) = .478, p = .033.

Significant relationships between physical barriers and the percentage of evasive vehicle actions
were also found in two types of pedestrian crossings. First, a significant relationship was found
between physical barriers and the percent of vehicle evasive actions to pedestrians crossing only
during the don’t walk phase in the marked intersection, rs(18) = .513, p = .021. A significant
relationship was found between physical barriers and vehicle evasive actions while pedestrians
crossed in the unmarked non-intersection with traffic, rs(18) = .513, p = .021.

Several significant relationships were found between categorical environmental factors and the
percentage of vehicle evasive actions in each crossing type. A significant relationship was found
between vehicle evasive actions and locations with a dedicated right turn only lane both while
pedestrians crossed entirely in the walk phase in the marked intersection, rpb(18) = .527,
p = .017, and while pedestrians crossed in the marked intersection overall, rpb(18) = .508,
p = .022.

Evasive Action Comparison

It is important to understand overall evasive action behavior. Here, pedestrian and vehicle
evasive actions are compared. First, evasive actions behaviors within the marked intersection
were compared. A significantly greater percentage of crossings involved pedestrian evasive
actions (M = 2.16 percent) than vehicle evasive actions (M = 0.04 percent), t(19) = -4.70,
p < .001. Next, evasive actions in unmarked non-intersections were examined. A significantly
greater percentage of crossings involved pedestrian evasive actions (M = 7.47 percent) than
vehicle evasive actions (M = 0.21 percent), t(19) = 4.65, p < .001.

Next, rule-following was examined. When pedestrians crossed in the marked intersection
entirely during the walk phase, significantly more pedestrians took evasive actions
(M = 1.02 percent) than did vehicles (M = 0.04 percent), t(19) = 2.96, p = .008. When
pedestrians made a rule-breaking crossing, there was a significant difference in the percentage of
vehicles that took evasive actions (M = 0.19 percent) and pedestrians that took evasive actions
(M = 8.91 percent), t(19) = 4.97, p < .001.

DISCUSSION

The goal of this work was to explore the factors that influence pedestrians’ roadway crossing
locations. Factors intrinsic to both humans and the environment were explored. The following
describes the findings from literature and phases 1, 2, and 3 of the present study.

Pedestrian Factors

A literature review revealed five traits and factors that influence pedestrians’ decisions on when
and where to cross the roadway: age, gender, alcohol, self-identity as a safe person, and
perceived control. While these influences are intertwined, each is discussed individually here.

190
Age not only plays a role in the decision on when and where pedestrians might choose to cross
the roadway, but also in the outcome of a collision. Children under the age of 15 make up
25 percent of pedestrian injuries, a value larger than any other age group.(7) Given this fact, it
may be worthwhile to target special interventions and educational tools to this age group to
reduce pedestrian–vehicle collisions. For example, in areas near schools, crosswalk activation
buttons may need to be lowered to child level or made visually more attractive to push. Design
modifications, coupled with educational tools that incorporate best crossing practices, may be
effective in reducing child pedestrian–vehicle collisions.

On the opposite end of the age spectrum, older adults make up about 19 percent of all pedestrian
fatalities, but only 8.5 percent of the total injuries.(7) Further, many studies have found that older
adults are less likely to attempt dangerous or potentially dangerous crossings.(8,9,10) These data
combined suggest that while older adults are less likely to be involved in a pedestrian–vehicle
crash, the outcome is likely to be more severe in terms of pedestrian health impact. As such, the
safety countermeasures that may be the most beneficial for older adults may not be effective in
all areas. For example, in areas near retirement communities, it may be useful to provide for
longer crossing times in marked crosswalks or include an opportunity to request (via push button
or other movement technology) a longer light protected crossing.

Pedestrian gender has both been associated with both risk crossing and higher pedestrian fatality
rates.(6,7,8) These findings are consistent with the general body of literature suggesting that males
are more likely to make risky decisions than females.(41) Given the inherent difficultly in
designing crosswalks and safety interventions specifically for one gender, time may be better
spent on educational interventions. In other words, specific advertisement and educational
campaigns can be used to help males better understand crossing risks and the best way to share
the roadway as a pedestrian.

As was previously noted, alcohol is involved in a relatively large proportion of fatal vehicle–
pedestrian crashes. Obviously, it is not legal to drive while intoxicated. However, walking is
generally thought be to be a safe mode of transportation. It is difficult to design pedestrian
crossing countermeasures specifically for the intoxicated pedestrian. However, measures can be
taken to reduce the fatalities in areas where pedestrians are likely to be intoxicated (e.g., bar
districts). For example, railings or shrubbery that separates the sidewalk from vehicular traffic,
and road closures might be used. However, these are not likely to be feasible on a widespread
basis. As a result, other tactics, such as educational information, public service announcements,
or targeted jay-walking enforcement, may reduce less safe pedestrian crossings.

Those people who identify themselves as safe pedestrians are less likely to make risky crossings
and accept fewer vehicle gaps as safe for crossing.(10) However, it seems unlikely that one’s self-
identity can easily be manipulated. In other words, this is likely a stable characteristic that is not
easily changed.

Finally, perceived control has also been shown to influence when and where pedestrians are
willing to cross the roadway.(9) When pedestrians perceive more behavioral control of the
situation, they are more likely to cross (or intend to cross) the roadway. As a result, it is logical
that roadways that present less perceived control (or predictability) to pedestrians are less likely
to have a high proportion of crossings outside the marked intersection during the walk light

191
phase (where, presumably, control would be the greatest). Along the same lines then, it may be
possible to modify roadway design to reduce perceived control away from marked intersections
and increase perceived control near marked intersections through affordance modifications. For
example, a median might increase control because it allows pedestrians to break up crossings,
completing one segment, seeking refuge, and completing the crossing when ready. Another way
to increase control is to allow pedestrians to activate walk signals via push button (with feedback
of activation initiation). However, one must also account for unintended consequences of
affordance modification. For example, if barriers are placed between the sidewalk and roadway,
people parking vehicles may not be able to access the sidewalk and may be forced to walk
unnecessarily close to vehicular traffic.

Environmental Factors

The remainder of this study focused on which specific environmental factors influence where
and when pedestrians cross the roadway. Specifically, pedestrian crossings at 20 different
locations in the Washington, DC, metropolitan area were recorded. The area of the crossing
(marked crossing or unmarked non-intersection crossing), timing of the crossing (with traffic or
against traffic), and other crossing circumstances were recorded (e.g., yielding and evasive
actions). For these 20 locations, more than 70,000 crossings were coded. Two different
methodologies were used to collect these data. One methodology involved two researchers
manually counting and coding pedestrian crossings while observing people in vivo. One
advantage of this methodology is that the observers can move about and gain better vantage
points as needed. However, in many of the high pedestrian volume areas, it is extremely difficult
to obtain accurate counts of pedestrian crossings through in-person observations. To be better
able to more accurately code and flexibly collect pedestrian data, a second methodology of video
recording was used. This allowed the video to be coded in both slowed and speed motion to
maximize accuracy and efficiency. The use of the DDOT traffic management cameras provided a
simple and unobtrusive way to capture video. Along with the benefits of this data collection
methodology, there were also some challenges. The first challenge is that the primary use of the
cameras is for traffic management. As a result, the cameras can be moved at any time and may
not be capturing footage from the desired areas. Furthermore, there can be service disruptions
with the video equipment (e.g., recording failure, camera failure). Despite these issues, video-
based data collection proved to be both fruitful and reliable.

Prediction Models

Coded data from all 20 locations were combined to create two separate crossing prediction
models.

Model 1—Intersection Versus Non-Intersection Crossings:

The first model predicts which area of the road pedestrians will cross (marked intersection or
unmarked non-intersection) based on the features of the roadway environment. Although many
factors were examined, not all were incorporated into the model. The roadway environment
variables that were included in the model fell into three categories:

192
 1. Travel Pace and Phasing.
a. Length of the walk phase.
b. Length of the do not walk phase.
c. Travel pace.

2. Traffic Throughput.
a. AADT.
b. Traffic directionality (one- or two-way street).
c. Curb-to-curb distance (i.e., the width of the street).

3. Distance to Safety.
a. Distance to the next marked crosswalk.
b. Presence and type of median.
c. Presence of cross streets between marked crosswalks.

The mean probability of the model correctly predicting a crossing in the marked intersection was
.9079. In other words, across the 20 locations, the model correctly predicted an average of
90.79 percent of the crossings. The median prediction accuracy was 91.28 percent with a range
of 80.55 percent at Location 8 to 95.22 percent at Location 18. Overall, the model was able to
successfully predict crossing location (marked intersection versus unmarked non-intersection)
using the features of the roadway environment.

Model 2—Walk Versus All Other Crossings:

It is likely that many of the pedestrians observed making crossings who did not cross in the
unmarked non-intersections did not do so simply because the trip did not require a crossing in
that specific direction. In other words, pedestrians may not have made an east/west non-
intersection crossing because their travel origination and destination only required a north/south
path. As a result, general rule-breaking was also examined. Rule-breaking is any crossing that
does not take place entirely during the walk light phase in the marked intersection.

Although many factors were examined, not all were incorporated into the model. The roadway
environment variables that were included in the model fell into five categories:

1. Travel Pace and Phasing.

a. Length of the walk phase.
b. Length of the do not walk phase.
c. Travel pace

2. Traffic Throughput.
a. AADT.
b. Traffic directionality (one or two way street).
c. Curb-to-curb distance (i.e., the width of the street).

3. Distance to Safety.
a. Distance to the next marked crosswalk.
b. Presence and type of median.
c. Presence of cross streets between marked crosswalks.

193
 4. External Objects (barriers/vehicles) in Center of Road.
a. Presence of physical barriers that might prevent a pedestrian from crossing the
roadway.
b. Presence of a center turning lane.

5. External Objects (vehicles) on Sides of Road.

a. Presence of parking along the roadway.
b. Presence of a right turn only lane.
c. whether or not the far marked crosswalk is light controlled.

The mean probability of the model correctly predicting a rule-following crossing in the marked
intersection was .9079. In other words, across the 20 locations, the model correctly predicted an
average of 90.79 percent of the crossings. The median prediction accuracy was 91.28 percent
with a range of 80.55 percent at Location 8 to 95.22 percent at Location 18. This model was
more successful in predicting rule-following crossings in some locations than in others.
However, overall, the model was able to successfully predict rule-following crossings using the
features of the roadway environment.

Proportions of Crossing Types and Environmental Factors by Location

To better accommodate the different overall numbers of pedestrian crossings at each of the
crossing locations, a series of analyses were performed using proportions of crossings. These
proportions were used to examine both crossing type and environmental factors on crossings.

Crossing Location

Overall, the mean percentage of the crossings that took place in the marked intersection was
83.88. (This value was derived by calculating the percentage of crossings that took place in the
marked intersection at each location and then these values were taken together to calculate the
mean.) The range was 50.88 percent (at Location 3) to 96.98 percent (at Location 7) with a
median value of 88.5 percent.

Location 3 is an outlier in this group at 2.70 standard deviations below the mean. Further, when
its value is removed, the mean percentage of pedestrians crossing in the marked intersection
jumps to 85.61. Yet, when examining the percentage of pedestrians that crossed in unmarked
non-intersections, Location 3 does not stand out. In fact, pedestrians were approximately equally
likely to cross at the unmarked intersection as the marked intersection. This finding suggests that
pedestrians perceive the unmarked intersection to be a safe and acceptable place to cross the
roadway. This is likely the result of a few unique characteristics of Location 3. Between the two
marked intersections, there is a T-intersection that is light controlled for vehicular traffic (but not
for pedestrian traffic). The light phases allow pedestrians travelling to and from the side street to
begin crossing, wait on the median, and then complete the crossing when the vehicle light phase
changes or an acceptable gap is presented. Given that 41 percent of the crossings at the
unmarked intersection involved waiting on the median, it appears that pedestrians were not
trapped on the median as previous research has reported. Rather, at this location, pedestrians plan
their crossing in phases—crossing a segment, waiting on the median, and then completing the
crossing. This is a tactic that presumably increases perceived control. Beyond the ability to

194
divide the crossing into two portions, there are environmental factors that both encourage
crossing at the unmarked intersection and discourage traveling to the marked intersection.

This path is likely desirable to pedestrians for several reasons. First, the path to/from the
neighboring Metro stop (about 1 block west) and the neighborhood area on the side street is more
direct when crossing at the unmarked intersection. Because Rhode Island Avenue Northwest is a
diagonal street, pedestrians are required to travel further south and then back north if they elect
to cross at the marked intersection. In other words, more travel distance is required. Furthermore,
the marked crossing is multipart with variant timing, which means that pedestrians may need to
wait on a median/refuge island to complete a crossing, independent of which intersection is used
for crossing in this area. To pedestrians, these two factors combined may outweigh the potential
benefits of crossing during a protected light phase—especially given the relatively rare
occurrence of a vehicle–pedestrian collision. As a result, pedestrians are likely to select to cross
at the unmarked intersection as a result of both convenience and a likely perception of more
control. Beyond this, the unmarked intersection has several factors that afford a crossing. It is
obvious that this is at an intersection; locations where most marked crossings take place. Further,
this lies at a junction where it is natural to want to travel. There is Metro station only one block
west of the marked intersection, on the north side. As a result, crossing at the unmarked
intersection when traveling to/from the Metro station, along Marion Street, is likely the most
direct and efficient route. Finally, to the pedestrian, the median in the unmarked intersection
looks as though it is a sidewalk (see figure 8). Pedestrians can clearly see a concrete area on the
end of the median that is approximately the width of a standard sidewalk (e.g., a firm, raised
surface that serves as a barrier from roadway vehicles). As such, this area affords the same things
to pedestrians as a standard sidewalk. It is likely that pedestrians interpret this area as they would
any other sidewalk area.

Based on these findings, many recommendations can be made for the design of new roadways or
the implementation of roadway environment modifications if crossing location (and timing) is a
consideration. First, it appears that pedestrians will treat intersections (whether marked with a
crosswalk or not) as an acceptable place to cross the roadway. Given sufficient pedestrian
volume, protected pedestrian crossings should be taken into account when designing
intersections. This is especially important if the intersection is in route to/from a high-density trip
originator.

A mean percentage of 13.89 of crossings occurred in unmarked non-intersection areas. This

value is especially low considering the percentage of pedestrian fatalities that occur at non-
intersection locations (approximately 64 percent). The discrepancy between these two
percentages could be the result of several factors. The first possibility is that pedestrians in the
Washington, DC, metropolitan area are fundamentally different than other pedestrians. However,
this seems unlikely. Another possibility is that because pedestrians infrequently cross outside
intersections, drivers do not anticipate pedestrians in these areas. Consequently, drivers are not
able to react quickly enough to avoid collision. Furthermore, traffic speeds are most often higher
at non-intersections than intersections, thus the outcomes of collisions are more severe.

Rule-following and rule-breaking crossings were explored. A mean percentage of

70.89 pedestrians crossed entirely during the walk phase of the marked intersection. That is,
most pedestrians do complete rule-following crossings. Location 13 was an outlier with only

195
28.41 percent of the crossings completed being rule following. This value is not surprising
considering the travel pace required to cross during the walk phase. The walk signal is only
illuminated for 10 s and the roadway is approximately 50 ft wide. This requires pedestrians to
travel at a rate of 5 ft/s, which is outside the range of comfortable travel for many people. This
rate is greater than the MUTCD recommended rate of 3.5 ft/s. Furthermore, the MUTCD states:
“Where pedestrians who walk slower than 3.5 feet per second, or pedestrians who use
wheelchairs, routinely use the crosswalk, a walking speed of less than 3.5 feet per second should
be considered in determining the pedestrian clearance time.” If one is concerned with rule-
breaking crossings, the length of the walk phase should be long enough to allow most pedestrians
to cross at a reasonable pace and at minimum, meet the MUTCD standards for time provided to
cross the roadway.

Analyses also revealed relationships between environmental factors and where pedestrians
crossed the roadway. First, a relationship between the width of the crossing and rule-following
was found. The wider the crossing, the more likely pedestrians are to make a rule-following
crossing. Or, alternatively, pedestrians are more likely to make a rule-breaking crossing when the
road is narrow. In addition, the presence of physical barriers along the sidewalk influenced where
pedestrians crossed the roadway. Specifically, the more objects that prevent easy access from the
sidewalk to the roadway, the less likely pedestrians were to cross at unmarked non-intersection
locations. Given these findings, for areas where there might be a problem with dangerous non-
intersection crossings, a simple intervention of increasing the barriers along the sidewalk could
be made. Placing obstacles like flower planter beds, benches, or other decorative barriers will not
eliminate the ability to cross the roadway. However, these barriers will reduce the affordance to
cross outside the marked intersection (and maybe even increase the visual esthetic of the area).

Somewhat interestingly, traffic directionality was significantly negatively correlated with

crossings made in the marked intersection entirely during the don’t walk phase. In other words,
pedestrians were more likely to cross entirely during the don’t walk light phase signal on one-
way streets than on two-way streets. This finding makes sense given that pedestrians are better
able to predict traffic that is approaching from one direction rather than two. While changing
roadway directionality may not be an appropriate roadway safety intervention, it should be
considered when designing/modifying existing road segments. For example, if it is known that
pedestrians are more likely to cross a one-way street during the don’t walk phase, the speed
could be modified to reduce the impact of potential conflicts or pedestrians could be provided
with an activation button that triggers the walk phase (increasing perceived control of the
crossing).

Pedestrian Yielding:

Of the total crossings at each location, there was a mean percentage of .98 that involved
pedestrian yielding. Although pedestrian yielding is discussed, it should not be taken as an
indication that there was potential for a collision. Rather, it was simply an action taken by
pedestrians to allow vehicles to pass. In the marked intersections, the mean percentage of
pedestrians that yielded to vehicles was .41. Similarly, the mean percentage of pedestrians
yielding to vehicles during rule-following crossings was .43. Location 17 was an outlier in both
cases, with 6.52 percent of the crossings in the marked intersection involving a pedestrian
yielding to a vehicle. All of these yielding behaviors were observed while pedestrians crossed

196
entirely during the walk phase, resulting in yielding in 7.27 percent of the rule-following
crossings. The exact circumstances of these yielding behaviors were not recorded. However,
observational notations indicate that these actions are the result of pedestrians yielding to
vehicles turning on to and from the busier main street (North Washington). In other words,
pedestrians allowed vehicles to pass, which increased vehicular traffic flow. While patterns such
as these show common roadway courtesy and may increase the overall level of user satisfaction,
they may lead to unforeseen consequences. For example, if a group of pedestrians yields to a
vehicle, the vehicle may proceed to turn to cross the path of the pedestrians. This is not an issue
if all pedestrians yield. However, an incident may occur if another pedestrian does not yield to
the driver, who is assuming that all pedestrians are yielding.

Overall, Location 10 was an outlier with 13.25 percent rule-breaking crossings involving
yielding to a vehicle. Additional notations indicate that some of these yielding behaviors may
have been the result of a person parking on one side of the street and seeking a trip destination on
the opposite side of the street. That is, pedestrians began crossing somewhere between the two
marked intersections and ultimately yielded to a vehicle before completing the crossing. If
parking away from a marked intersection does indeed lead people to cross at unmarked non-
intersection crossings to reach destinations on the opposite side of the road, it may be difficult to
implement interventions to promote crosswalk use. However, items such as barriers that create
separation between bi-directional traffic and reduce crossing affordances may be effective
(although they have many drawbacks, including cost and feasibility).

As one might expect, there were significantly more instances of pedestrians yielding to vehicles
in unmarked non-intersection areas than in marked intersections. Similarly, there were
significantly more pedestrian yielding behaviors in rule-breaking crossings than rule-following
crossings. These findings suggest that when pedestrians do make rule-breaking crossings, they
are prepared to yield to vehicles.

Pedestrian yielding is also significantly correlated with several environmental factors. There was
a significant correlation between the length of the walk signal light phase and the percentage of
pedestrians that yielded to vehicles in the marked intersection. This is perhaps not surprising. If
one thinks about a moderate to heavy volume crossing, pedestrians often group together as they
wait for the signal to change. After the initial bulk of pedestrian traffic has crossed the
intersection, queued vehicles begin to make turns onto perpendicular streets. As pedestrians, who
had not been waiting to cross, reach the intersection, they may yield to allow the vehicles to
complete crossing. This serves both to keep vehicular traffic flowing and to provide pedestrians
with some control of their safety in the crossing.

Traffic volume, as estimated by AADT, was also significantly correlated with pedestrians
yielding during rule-breaking crossings. This is not surprising given that as vehicular traffic
increases, there are more opportunities for pedestrians to yield to vehicles. Given these data, it is
difficult to say that areas with higher AADT are necessarily more dangerous for pedestrians.
Rather, pedestrians are simply more likely to yield to vehicles, which may actually indicate a
safer area for pedestrians.

197
Vehicle Yielding:

Overall, an average of 8.93 pedestrian crossings involved a vehicle yielding to a pedestrian.

Location 7 was an outlier, with 38.69 percent of the crossings involving vehicle yielding. This
trend remained consistent with both crossings in the marked crossing (39.80 percent) and rule-
following crossings (50.71 percent). This area had a substantial number of vehicles turning right
and passing through the intersection of interest. As a result of the turning and through traffic
having a green light (but not a protected right turn) and the pedestrian signal in the walk phase,
vehicles often waited to complete their turn (i.e., yielded) while pedestrians crossed the roadway.
These yielding behaviors do not necessarily indicate a safety concern. However, turning traffic in
high pedestrian volume areas should be evaluated in intersection design.

Overall, vehicles yielded to 2.55 percent of pedestrians making rule-breaking crossings. Both
Location 17 (11.67 percent) and Location 12 (9.23 percent) were outliers. Location 12 is unusual
because there were very few rule-breaking crossings overall. As a result, the six observed vehicle
yielding behaviors resulted in a high percentage of the total rule-breaking crossings. Location 17
however, is unusual in a different way. This block is in an area known as Old Town Alexandria.
In this particular neighborhood, the street is relatively narrow, has parking on both sides of the
street, has slow-moving traffic (often looking for parking), and is surrounded by small shops and
businesses. This is an area where people move at a leisurely pace. These factors combine to
create a unique atmosphere where drivers often stop and “wave” pedestrians to cross the road.
Although these specific actions were not measured, they were observed and noted on many
crossings. This vehicle yielding in this area may suggest that cultural differences in
neighborhood may influence the need (or lack of a need) for different safety interventions.

Overall, significantly greater percentages of vehicles yielded to pedestrians crossing in the

marked crossing than in the unmarked non-intersection. The same pattern is true with rule-
following and rule-breaking crossings.

A relationship between the number of trip originators and destinations in an area and vehicles
yielding to pedestrians making rule-breaking crossings was found. That is, the more trip
originators and destinations present, the higher the percentage of pedestrian rule-breaking
crossings that involved vehicle yielding. These findings suggest pedestrians may be more
motivated to make potentially dangerous rule-breaking crossings to reach a trip destination. It
also points out the necessity to take special consideration of pedestrian crossing facilities in areas
with a high density of commercial businesses.

Yielding Comparison:

Not surprisingly a significantly greater percentage of vehicles than pedestrians yielded in rule-
following crossings. The same is true for crossings taking place in the marked intersection
overall. These findings suggest that drivers are respectful of shared road use with pedestrians in
the marked intersections. However, it is not known whether this relationship remains true for
non-intersection marked crossings (i.e., not light-controlled marked midblock crossings). This
specific type of crossing was not explored in the present study. However, vehicles and
pedestrians were equally likely to yield in unmarked non-intersection areas. The same remains
true in rule-breaking crossings overall.

198
It is perhaps a bit surprising that there were no significant differences in vehicle and pedestrian
yielding outside the marked intersections and in rule-breaking crossings. It might be expected
that pedestrians would take a more proactive safety approach when making a rule-breaking
crossing. The lack a difference in yielding may also provide insight into the disproportionate
percentage of crossings at non-intersection locations and high fatality rates in these areas. This
may also indicate that vehicles will yield proactively to allow pedestrians to cross outside the
normal rule following crossings and that pedestrians expect vehicles to yield. Both have
implications for pedestrian safety and potential collisions.

Evasive Pedestrian Actions:

The mean percentage of crossings that included an evasive pedestrian action was 3.06. Both
Location 1 and Location 8 were outliers. Location 8 requires pedestrians to travel at a pace of
3.7 ft/s, faster than the recommended rate of 3.5 ft/s. Given that there is an elementary school
approximately two blocks away from this intersection and that there are other small suburban
type establishments (including a church, neighborhood market, and library) in the general area, it
is likely that there is substantial pedestrian traffic near this intersection that regularly travels at a
rate of less than 3.5 ft/s.

While pedestrians were not specifically queried about their crossings, it is possible that
pedestrians may feel rushed while crossing at the marked intersection. This rapid pace required
to cross at the marked intersection during the walk phase may lead pedestrians to feel hurried and
uncomfortable crossing at this location. Furthermore, if pedestrians are not able to complete the
crossing during the walk phase, they may be forced to take an evasive action to complete the
crossing. This intersection is also just outside a traffic circle. This prevents pedestrians from
being able to view traffic from a distance adequate to determine whether it will continue
traveling within the traffic circle or exit toward the intersection. Furthermore, this reduces
pedestrians’ abilities to confidently determine whether the vehicle will cross their potential path
during a crossing. At this location, pedestrians may begin a crossing but not be able to clearly
view all vehicular traffic. As a result, pedestrians might be forced to take evasive actions to
complete the crossing safely.

Location 1 was an outlier both in terms of evasive actions in marked intersections and unmarked
non-intersections. At Location 1, there were many instances of “courtesy” acceleration by
pedestrians. In other words, it appeared that pedestrians would run, or accelerate, through a
crossing to allow turning vehicles to complete their turn during the signal phase. This turning
traffic may have also been the cause of the evasive actions made outside the marked intersection.
Pedestrians may have started crossing the roadway while no traffic was visible. However, after
the pedestrian initiated the crossing, a vehicle could have turned onto the main roadway and
forced the pedestrians to take an evasive action to avoid a potential collision. These
circumstances were not specifically coded in the data; however, several notes indicated that this
did in fact occur. It may not be possible to target a specific intervention for these potential
sources of conflict. However, vehicle turning phases (or lack thereof) and pedestrian density
should be considered in roadway design.

As one might expect, pedestrians had a significantly greater percentage of crossings in unmarked
non-intersection areas that involved an evasive pedestrian action than crossings in the marked

199
intersection. This same pattern is true for rule-following and rule-breaking crossings. In other
words, when pedestrians make a rule-breaking crossing, they are more likely to need to take an
evasive action to avoid collision.

A significant negative relationship between the percentage of evasive actions taken in unmarked
non-intersections and the length of the walk light phase was found. That is, the shorter the walk
phase, the more likely that crossings in the unmarked non-intersection will involve an evasive
pedestrian action. Thus it appears that light phasing may be an important factor to consider in
terms of potential vehicle/pedestrian conflicts. However, light phase timing is also associated
with traffic density. As a result, simply increasing the length of the walk signal is not likely to
eliminate all pedestrian evasive actions.

Evasive Vehicle Actions:

Overall, the mean percentage of pedestrian crossings that involved an evasive vehicle action was
.10. Location 12 was considered an outlier in the percentage of crossings involving an evasive
action, both overall and unmarked non-intersection areas. However, only two vehicle evasive
actions were recorded at this location. Given the small number, these crossings are not discussed
further.

Location 17 is also an outlier in terms of the percentage of evasive vehicle actions made during
rule-following crossings and crossings in the marked intersection overall. However, Location 17
only had one recorded evasive vehicle action. As a result, it is not discussed further here.

No significant difference between the percentage of crossings in the marked intersection and the
unmarked non-intersection that included a vehicle evasive action was found. No difference in the
percentages between rule-following and rule-breaking crossings was found either.

A significant correlation between the length of the walk phase and the percentage of evasive
vehicle actions during marked intersection pedestrian crossings was found. In other words, the
longer the walk phase, the more likely vehicles were to take an evasive action while pedestrians
crossed in the marked intersection. Given the relationship between pedestrian yielding in the
marked intersection and the length of the walk phase, this may not be surprising. As discussed in
a previous example, it is simple to imagine how pedestrians might group together to wait for a
signal to change at a moderate to heavy volume crossing area. After the initial bulk of pedestrian
traffic has crossed the intersection, queued vehicles begin to make turns onto perpendicular
streets, passing through the marked pedestrian crossing. However, pedestrians who reach the
intersection later may come as a surprise to vehicles. As a result, drivers may be forced to make
an evasive maneuver to avoid collision.

Evasive Action Comparison:

For crossings made both in the marked intersection and the unmarked non-intersection,
pedestrians had a significantly greater percentage of evasive actions than did vehicles. The same
pattern is true for rule-following and rule-breaking crossings. This difference can be interpreted
in many different ways. In general, it appears as though pedestrians proactively take action to
avoid potential collisions. That is, independently of whether pedestrians are making a rule-
following or a rule-breaking crossing, they are more likely to take an evasive action to avoid

200
collision than are drivers. Challenges may arise, however, when pedestrians are not able to take
an appropriate evasive action, are not aware that one is needed, or when a driver anticipates a
pedestrian will take an evasive action and he or she does not.

Summary

The overarching goal of this study was to determine which factors influence when and where
pedestrians cross the roadway. Both factors associated with pedestrians and factors associated
with the crossing environment were examined. Although factors intrinsic to pedestrians are not
easily manipulated, measures that increase safety can be implemented in areas dominated by one
specific demographic group (e.g., older adults near a retirement community or smaller children
near a school). Furthermore, educational methods can be used to help pedestrians make wiser
crossing choices.

Environmental factors were explored at 20 different crossing locations. A model was developed
that relatively successfully predicted whether pedestrians would cross within the marked
crossing or not. While this model is not 100-percent accurate, it does provide a basis to
understand where pedestrians might cross the roadway. This model may prove useful in the
design, or renovation, of new roadways and shared use communities. Areas that have a high
predicted likelihood of unmarked non-intersection crossings could be proactively targeted to
modify the crossing affordances of the environment.

The data here provide insights into some of the specific environmental factors that influence
where and when pedestrians cross the road. Some suggestions are provided for ways these
factors can be modified to influence pedestrian behavior (e.g., reducing crossing affordances by
adding flower beds separating the sidewalk and roadway). Evidence is also provided that
suggests modifying environmental affordances to alter perceived control of the crossing will
modify where pedestrians cross the roadway. This can be accomplished by both increasing the
control of the marked crossing signal (e.g., activated button presses, timed countdowns, traffic
flow information) and reducing control outside the marked crossings (e.g., including road
barriers, removing medians in bi-directional traffic). However, it should be noted that these
modifications can have unintended consequences in crossing behavior and should be evaluated
carefully. The exact interaction and interplay of these factors is not fully understood.

An important consideration in environmental factor manipulation is the safety impacts of

modifying where pedestrians cross the roadway. It is not known whether pedestrian crossings in
the intersection are actually safer than crossings made at non-intersection locations. In fact, of
the crossings in the 20 different locations, the only near incident took place in a marked
intersection. At Location 7, two pedestrians were about to enter the roadway while at the same
time, a vehicle began to make a right turn (passing through the marked intersection). One person
pulled the other person back to the curb. Although it is not clear from the video, it appears that
the vehicle may have nearly collided with the pedestrian had he or she not be pulled back to the
curb. Despite the pedestrian having the right of way in this case, it is easy to understand why the
driver continued with the right turn. Prior to the pedestrians attempting to cross the roadway,
they stood on the curb facing the opposite crossing direction (i.e., facing north, rather than east).
Furthermore, the walk phase had been initiated for more than 10 s prior to the pedestrians
attempting to enter the marked intersection. These clues could easily lead a driver not to interpret

201
the pedestrians as a potential hazard. Therefore, it does appear that predictability of pedestrian
actions plays a key in driver reaction and potential safety.

In sum, these data help to explain where, when, and why pedestrians cross the road. Through this
report, the data and analysis help to promote a better understanding of pedestrian behavior. This
information can be used to evaluate the implementation of new roadways and communities,
redesign of existing environments, and identification of areas that may benefit from engineering
interventions such as crossing-inhibiting designs or pedestrian hybrid beacons.

202
 REFERENCES

1. Fatality Reporting Analysis System (2011). http://www-

fars.nhtsa.dot.gov//QueryTool/QuerySection/SelectYear.aspx Accessed November 2011.
2. Jacobs, G., Aeron-Thomas, A., and Astrop, A. (2000). Estimating Global Road Fatalities,
TRL Report 445, TRL Ltd., Crowthorne, UK.
3. Naci, H., Chisholm, D., and Baker, T.D. (2009). “Distribution of Road Traffic Deaths by
Road User Group: A Global Comparison,” Injury Prevention, 15, 55–59.
4. World Health Organization (2004). World Report on Road Traffic Injury Prevention.
5. Karkee, G.J., Pulugurtha, S.S., and Nambisan, S.S. (2009). Statistical Analysis of Pedestrian
Crossing Behavior on Streets. Presented at 88th Annual Meeting of the Transportation
Research Board, Washington, DC.
6. National Highway Transportation Safety Administration (NHTSA) (2011). Traffic Safety
Facts, 2009, U.S. Department of Transportation, Washington, DC.
7. NHTSA (2011). Traffic Safety Facts: Pedestrians. U.S. Department of Transportation,
Washington, DC.
8. Holland, C., Hill, R. (2007). “The Effect of Age, Gender, and Driver Status on Pedestrians’
Intentions to Cross the Road in Risky Situations,” Accident Analysis and Prevention, 39,
224–237.
9. Evans, D. and Norman, P. (1998). “Understanding Pedestrians' Road Crossing Decisions: an
Application of the Theory of Planned Behavior,” Health Education Research, Theory and
Practice, 13, 481–489.
10. Holland, C.A., Hill, R., and Cooke, R. (2009). “Understanding the Role of Self-Identity in
Habitual Risky Behaviours: Pedestrian Road-Crossing Decisions Across the Lifespan,”
Health Education Research, 24(4), 674–685.
11. Barton, B.K. and Morrongiello, B.A. (2011). “Examining the Impact of Traffic Environment
and Executive Functioning on Children’s Pedestrian Behaviors,” Developmental Psychology,
47, 182–191.
12. Evans, D. and Norman, P. (2003). “Predicting Adolescent Pedestrians’ Road-Crossing
Intentions: An Application and Extension of the Theory of Planned Behaviour,” Health
Education Research, 18(3), 267–277.
13. Conner, M. and Sparks, P. (1996). “Theory of Planned Behavior and Health Behaviour,” in
Conner, M. and Norman, P. (Eds), Predicting Health Behaviour. Open University Press,
Buckingham, 121–162.
14. Gibson, J.J. (1966). The Senses Considered as Perceptual Systems, Houghton Mifflin,
Oxford.
15. Tyrrell, R.A., Wood, J.M., Carberry, T.P. (2004). “On-Road Measures of Pedestrians’
Estimates of Their Own Nighttime Conspicuity,” Journal of Safety Research, 35,
483–490.
16. Norman, D.A. (1988). The Psychology of Everyday Things, Basic Books, New York.
17. Retting, R.A, Ferguson, S.A., and McCartt, A.T. (2003). “A Review of Evidence-Based
Traffic Engineering Measures Designed to Reduce Pedestrian-Motor Vehicle Crashes,”
American Journal of Public Health, 9, 1,456–1,463.
18. District of Columbia Department of Transportation (2012). Annual Average Daily Traffic,
2010 Traffic Volume Map.
http://ddot.dc.gov/DC/DDOT/About+DDOT/Maps/Traffic+Volume+Maps

203
19. Google®, Inc. (2012) Map of the intersection of H and 3rd Streets Northeast in Washington,
DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b89bc09737f4f2f2
e&msa=0&ll=38.900318,-
77.001328&spn=0.016065,0.026071&iwloc=0004b89bc94bb3a1299df)
20. Google®, Inc. (2012) Map of the intersection of 7th Street and New York Avenue Northwest
in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b89bc09737f4f2f2
e&msa=0&ll=38.9029,-
77.021222&spn=0.002008,0.003259&iwloc=0004b89bcce2495d74a9b)
21. Google®, Inc. (2012) Map of the intersection of 7th Street and Rhode Island Avenue
Northwest in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b89bc09737f4f2f2
e&msa=0&ll=38.912142,-
77.021354&spn=0.002008,0.003259&iwloc=0004b89bd078124769f02).
22. Google®, Inc. (2012) Map of the intersection of 13th and I Streets Northwest in Washington,
DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b89bc09737f4f2f2
e&msa=0&ll=38.90151,-
77.028786&spn=0.002008,0.003259&iwloc=0004b89bd40ef46163777)
23. Google®, Inc. (2012) Map of the intersection of 14th and H Streets Northwest in Washington,
DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b89bc09737f4f2f2
e&msa=0&ll=38.900137,-
77.031785&spn=0.002008,0.003259&iwloc=0004b89bd40f0865faecb)
24. Google®, Inc. (2012) Map of the intersection of 14th Street and New York Avenue Northwest
in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b89bc09737f4f2f2
e&msa=0&ll=38.899283,-
77.031927&spn=0.004016,0.006518&iwloc=0004b89bd7a3b7b5a3fee)
25. Google®, Inc. (2012) Map of the intersection of 19th Street, H Street, and Pennsylvania
Avenue Northwest in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b89bc09737f4f2f2
e&msa=0&ll=38.899283,-
77.040945&spn=0.004016,0.006518&iwloc=0004b89bd7a3cc098189f)
26. Google®, Inc. (2012) Map of the intersection of Oliver Street and Connecticut Avenue
Northwest in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b89bc09737f4f2f2
e&msa=0&ll=38.966412,-
77.075964&spn=0.004012,0.006518&iwloc=0004b89bdb3940c0107c2)
27. Cohen, R.A. (2009). “Applications of the GLMSELECT Procedure for Megamodel
Selection,” in Proceedings of the SAS Global Forum 2009 Conference, Cary, NC: SAS
Institute Inc.
28. Manual on Uniform Traffic Control Devices. FHWA, U.S. Department of Transportation,
Washington, DC, 2009.

204
29. Google®, Inc. (2012) Map of the intersection of Connecticut and Florida Avenues Northwest
in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b9b739239ddfdbc4
9&msa=0&iwloc=0004b9b74dba7f6a96117)
30. Google®, Inc. (2012) Map of the intersection of Connecticut Avenue and Van Ness Street
Northwest in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b9b739239ddfdbc4
9&msa=0&ll=38.943164,-
77.062901&spn=0.008027,0.013036&iwloc=0004b9b7514fc6ad1e1fb)
31. Google®, Inc. (2012) Map of the intersection of Georgia and Arkansas Avenues Northwest in
Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b9b739239ddfdbc4
9&msa=0&ll=38.951057,-
77.027088&spn=0.004013,0.006518&iwloc=0004b9b758797dd2cd556)
32. Google®, Inc. (2012) Map of the intersection of Georgia Avenue and Irving Street Northwest
in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b9b739239ddfdbc4
9&msa=0&ll=38.928956,-
77.023065&spn=0.004015,0.006518&iwloc=0004b9b766c8b281caacc)
33. Google®, Inc. (2012) Map of the intersection of H Street and Connecticut Avenue Northwest
in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b9b739239ddfdbc4
9&msa=0&ll=38.899792,-
77.03639&spn=0.004016,0.006518&iwloc=0004b9b78366bd4822ba7)
34. Google®, Inc. (2012) Map of the intersection of Independence and Washington Avenues
Southwest in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b9b739239ddfdbc4
9&msa=0&ll=38.887534,-
77.013184&spn=0.004017,0.006518&iwloc=0004b9b78a8fe40fe0aed)
35. Google®, Inc. (2012) Map of the intersection of Pennsylvania Avenue and 8th Street
Southeast in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps/ms?msid=207498904521634712066.0004b9b739239ddfdbc4
9&msa=0&ll=38.884348,-
76.99517&spn=0.004017,0.006518&iwloc=0004b9b744140e0168e29)
36. Virginia Department of Transportation (2012). Annual Average Daily Traffic, 2010 Traffic
Volume Map. http://www.virginiadot.org/info/2010_traffic_data_by_jurisdiction.asp
37. Google®, Inc. (2012) Map of the intersection of King and North Washington Streets in
Alexandria, VA. Accessed September 19, 2012
(https://maps.google.com/maps?q=king+and+north+washington,+alexandria,+va&hl=en&ll=
38.804868,-77.046561&spn=0.004022,0.006518&sll=38.925434,-
77.035778&sspn=0.128472,0.208569&t=h&hnear=King+St+%26+N+Washington+St,+Alex
andria,+Virginia+22314&z=18)

205
38. Google®, Inc. (2012) Map of the intersection of Irving and 14th Streets Northwest in
Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps?q=irving+and+14th+Washington+dc&hl=en&sll=
38.804868,-77.046561&sspn=0.004022,0.006518&t=h&hnear=
14th+St+NW+%26+Irving+St+NW,+Washington,+District+of+Columbia&z=17)
39. Google®, Inc. (2012) Map of the intersection of Wisconsin and Western Avenues Northwest
in Washington, DC. Accessed September 19, 2012
(https://maps.google.com/maps?q=willard+ave+and+wisconsin+ave+washington+dc&hl=en
&ll=38.959943,-77.085571&spn=0.008026,0.013036&sll=38.928774,-
77.03264&sspn=0.008029,0.013036&t=h&hnear=Wisconsin+Ave+%26+Willard+Ave,+7,+
Montgomery,+Maryland+20815&z=17)
40. Google®, Inc. (2012) Map of the intersection of Clarendon Boulevard and North Edgewood
Street in Arlington, VA. Accessed September 19, 2012
(https://maps.google.com/maps?q=North+Fillmore+Street,+Clarendon,+Arlington,+VA&hl=
en&ll=38.888127,-77.09188&spn=0.004017,0.006518&sll=38.959943,-
77.085571&sspn=0.008026,0.013036&oq=clarendon+north+fillmore&t=h&hnear=N+Fillmo
re+St,+Arlington,+Virginia+22201&z=18)
41. Wilson, M. and Daly, M. (1985). “Competitiveness, Risk Taking, and Violence: The Young
Male Syndrome,” Ethology and Sociobiology, 6, 59–73.

206
HRDS-30/01-14(WEB)E