Potentially Replaceable Car Trips: Assessment of Potential Modal Change towards Active Transport Modes in Vitoria-Gasteiz

Abstract

Road traffic is the most important contributor to noise and air pollutant emissions in cities. Its substitution by non-motorized modes therefore has great potential to improve the urban environment while increasing levels of physical activity among the population. This paper identifies car trips that could potentially be transferred to active modes such as walking and cycling, and analyses the barriers perceived by people who travel by car. We detect potentially replaceable car trips based on a mobility survey, distance calculation, and a distance threshold approach. The answers to a set of questions in the mobility survey allow us to identify the perceived barriers for use of the bicycle, applied to Vitoria-Gasteiz (Spain). The results show that between 30% and 40% of car trips could be replaced by active modes. Personal safety and distance results are the most limiting barriers perceived by car users, while physical condition and technique are the most limiting ones for bicycle users. These results provide valuable information for implementing measures to promote the replacement of motorized trips with walking and cycling.

Full text

sustainability
Article
Potentially Replaceable Car Trips: Assessment of
Potential Modal Change towards Active Transport
Modes in Vitoria-Gasteiz
Javier Delso 1, Belén Martín1, 2, * and Emilio Ortega 1,2
1Transport Research Centre (TRANSyT-UPM), Universidad Politécnica de Madrid, 28040 Madrid, Spain;
Javier.delso@upm.es (J.D.); e.ortega@upm.es (E.O.)
2Department of Forest and Environmental Engineering and Management, MONTES (School of Forest
Engineering and Natural Environment), Universidad Politécnica de Madrid, 28040 Madrid, Spain
*Correspondence: belen.martin@upm.es; Tel.: +34-91-0671646
Received: 29 August 2018; Accepted: 25 September 2018; Published: 30 September 2018


Abstract:
Road traffic is the most important contributor to noise and air pollutant emissions
in cities. Its substitution by non-motorized modes therefore has great potential to improve the
urban environment while increasing levels of physical activity among the population. This paper
identifies car trips that could potentially be transferred to active modes such as walking and cycling,
and analyses the barriers perceived by people who travel by car. We detect potentially replaceable car
trips based on a mobility survey, distance calculation, and a distance threshold approach. The answers
to a set of questions in the mobility survey allow us to identify the perceived barriers for use of
the bicycle, applied to Vitoria-Gasteiz (Spain). The results show that between 30% and 40% of car
trips could be replaced by active modes. Personal safety and distance results are the most limiting
barriers perceived by car users, while physical condition and technique are the most limiting ones for
bicycle users. These results provide valuable information for implementing measures to promote the
replacement of motorized trips with walking and cycling.
Keywords: active transport; transport modal change; barriers to cycling
1. Introduction
The second half of the 20th century saw the prioritization of motorized transport modes, which in
some cities came to occupy over 60% of the public space, with their associated and well-known
problems of noise, pollution, and safety [
1
]. Road traffic is the main contributor to noise and air
pollutant emissions in cities. It is estimated that between 16% and 25% of the European population
is exposed to pollution and noise levels above the established legal limits [
2
]. Exposure to these
emissions has a negative impact (physical and mental) on citizens’ health, including respiratory and
cardiovascular problems, anxiety, depression, and sleep disorders [
3
]. Up to 20% of the estimated
premature mortality related to urban traffic impacts on health could be avoided by implementing
new urban mobility policies [
4
]. One type of trip that is of particular interest in the urban planning of
sustainable transport is the short car trip (SCT), which accounts for a significant proportion of urban
motorized trips [
5
–
7
], and offers an opportunity for a modal shift towards more sustainable modes.
Engines consume more fuel and emit a greater ratio of pollutants [
8
,
9
] during SCT, so their substitution
by non-motorized modes has great potential to improve the urban environment while simultaneously
increasing the levels of physical activity among the population.
When considering the active modes that could replace SCT, one fact worth noting is that
walking is accessible to most of the population, and provides important health benefits [
10
]. However,
the distances that different individuals are willing to walk can vary depending on the trip purpose
Sustainability 2018,10, 3510; doi:10.3390/su10103510 www.mdpi.com/journal/sustainability

Sustainability 2018,10, 3510 2 of 13
(work, study, shopping, etc.) and other socio-economic characteristics (age, gender, occupation, etc.).
Previous studies have therefore used different thresholds to define short trips, differentiating between
population subgroups [
6
,
11
]. These studies consider trip chains—understood as the total distance
an individual travels in all the trips starting and ending in the same place [
12
]—and determine the
thresholds for substituting such trip chains made by car with active transport modes. For example, [
6
]
considers a single threshold of 6.4 km as a maximum for a walking trip chain, while [
11
] define this
distance using the 80th percentile of the total number of walked distances per day in a particular
survey. Previous studies have not always considered replacing the car with cycling as a possibility,
as the bicycle modal share was too low [
6
,
11
]. According to [
13
], replacing a car trip with a bicycle trip
is more complex than replacing a car trip with a walking trip. Nevertheless, due to the great potential
for replacing motorized trips with the bicycle, and the boom in this transport mode over the last few
years in parts of Europe, it is interesting to study cycling as a potential substitute for SCT.
The factors that can influence the modal selection of cycling include demographic characteristics
and socioeconomic variables such as personal income and education level [
13
,
14
], sex, and age,
among others [
15
,
16
]. Urban environment factors such as dedicated cycle lanes, green spaces,
population density, public transport, or urban safety can also act as facilitators or barriers to the use of
the bicycle [
17
–
20
]. Some studies in the literature have focused on investigating the differences between
perceptions of barriers among people who use the bicycle as a mode of transport and those who do
not [
14
,
21
]. Ref. [
22
] analysed the differences in a student sample between the barriers perceived
to walking and cycling to their study centre. It is important to note that when the trip distance
alone is considered as a barrier, a significant proportion of SCTs could be replaced by the bicycle.
The literature points to a number of limiting factors related to trip type, the availability of bicycles,
and infrastructures, all of which significantly reduce the SCT that are eligible for replacement [
7
,
19
].
Understanding the main perceived barriers to substituting these SCT with cycling offers a valuable
tool for transport planners, and can be used to design policy instruments to increase shared active
transport modes while decreasing motorized mobility.
This work studies the possible modal shift towards active transport modes, and has two objectives.
The first is to analyse the car trips that could potentially be transferred to active modes such as walking
and cycling. Once these trips have been identified, the second objective is to understand the barriers to
cycling perceived by people using SCT, focusing on a potential transfer towards active modes. This is
done by comparing the barriers to cycling between two groups of people: those who never use the
bicycle and those who use it occasionally. Vitoria-Gasteiz, a medium-sized city in northern Spain,
serves as a case study.
2. Data and Methods
2.1. Context and Data
The case study selected is Vitoria-Gasteiz, a city of 244,000 inhabitants in the province of Álava
(Spain). It is the capital of the Basque Country, and has evolved from an intensive industrial economic
model towards a more balanced one with the emergence of a service sector. Vitoria-Gasteiz is
characterized by its compactness and a clear separation between urban uses and the countryside
(see Figure 1). For over one decade, local authorities have worked on a sustainable transport policy
in the Mobility and Public Space Plan and the Cycling Mobility Transport Plan, which has led to an
increase in the use of active transport modes [23].

Sustainability 2018,10, 3510 3 of 13
Sustainability 2018, 10, x FOR PEER REVIEW 3 of 13
Figure 1. Map of Vitoria-Gasteiz.
This work is based on the 2014 Vitoria-Gasteiz household travel survey (HTS) [24]. This is a
telephone survey based on a stratified random sample containing data on trips made on the previous
day (origin, destination, mode, trip purpose, etc.) and socioeconomic data from 4192 individuals,
representing the total population of Vitoria-Gasteiz aged over nine years [24]. The HTS includes a set
of questions to describe the frequency of use and expertise of the respondents, and the perceived
limitations and barriers to the use of bicycle as means of transport [25,26]; nevertheless, it does not
address the perceived limitations and barriers to walking. The current study selects the trips by
individuals aged over 18 (the driving age), resulting in 3786 individuals and 16,037 trips. More
information about the contents and design of this survey can be found in [26].
The HTS was complemented with another data source. The Google Maps API [27] was used to
obtain the travel times for walking and cycling trips, using as origins and destinations the appropriate
set of origins and destinations from the HTS survey, as explained below.
2.2. Methods
The next two subsections describe the procedure followed to obtain the walking and cycling
travel times and distances that form the basis for calculating the thresholds to characterize the SCT
that can potentially be replaced by actives modes. They explain the statistical methods used to analyse
the barriers to cycling.
2.2.1. Analysing the Potential Replacement of SCT by Active Modes
The aim of the first part of the methodology is to identify the SCT that could potentially be
replaced by active modes using thresholds based in objective criteria: one for the walking mode and
three for the cycling mode. The first criterion for both modes is distance-based; it is defined using the
80th percentile of walking and cycling distances calculated for the sample. The second and third
criteria for the cycling mode are related to bike ownership and cycling experience of the respondents.
Regarding the first criterion, distance values are defined as boundaries (or thresholds) to identify
the car trips that could be shifted to sustainable modes; if the car trip is below these distance
thresholds, the car use can be considered as a candidate for substitution by walking or cycling.
Distances were calculated with the Google Maps API [27]. The origins and destinations of each HTS
Figure 1. Map of Vitoria-Gasteiz.
This work is based on the 2014 Vitoria-Gasteiz household travel survey (HTS) [
24
]. This is
a telephone survey based on a stratified random sample containing data on trips made on the
previous day (origin, destination, mode, trip purpose, etc.) and socioeconomic data from 4192
individuals, representing the total population of Vitoria-Gasteiz aged over nine years [
24
]. The HTS
includes a set of questions to describe the frequency of use and expertise of the respondents, and the
perceived limitations and barriers to the use of bicycle as means of transport [
25
,
26
]; nevertheless,
it does not address the perceived limitations and barriers to walking. The current study selects the
trips by individuals aged over 18 (the driving age), resulting in 3786 individuals and 16,037 trips.
More information about the contents and design of this survey can be found in [26].
The HTS was complemented with another data source. The Google Maps API [
27
] was used to
obtain the travel times for walking and cycling trips, using as origins and destinations the appropriate
set of origins and destinations from the HTS survey, as explained below.
2.2. Methods
The next two subsections describe the procedure followed to obtain the walking and cycling
travel times and distances that form the basis for calculating the thresholds to characterize the SCT
that can potentially be replaced by actives modes. They explain the statistical methods used to analyse
the barriers to cycling.
2.2.1. Analysing the Potential Replacement of SCT by Active Modes
The aim of the first part of the methodology is to identify the SCT that could potentially be
replaced by active modes using thresholds based in objective criteria: one for the walking mode and
three for the cycling mode. The first criterion for both modes is distance-based; it is defined using
the 80th percentile of walking and cycling distances calculated for the sample. The second and third
criteria for the cycling mode are related to bike ownership and cycling experience of the respondents.
Regarding the first criterion, distance values are defined as boundaries (or thresholds) to identify
the car trips that could be shifted to sustainable modes; if the car trip is below these distance thresholds,
the car use can be considered as a candidate for substitution by walking or cycling. Distances were
calculated with the Google Maps API [
27
]. The origins and destinations of each HTS trip are encoded

Sustainability 2018,10, 3510 4 of 13
and used as input for the API [
27
], which calculates the shortest distance for each trip and the main
transport mode declared in the survey. The Google Maps API [
27
] was also used to simulate car
trips by substituting the mode by cycling and walking to obtain the distance travelled in the case of
modal shifts.
After obtaining the distribution of the distances travelled in the sample, the thresholds are
calculated with a method based on the works of [
6
,
11
]. The distance thresholds for active travel
are established using the 80th percentile of walking and cycling distances calculated for the sample,
distinguishing between gender and age groups (18–34, 35–49, 50–64, and 65–84 years old). Thresholds
are calculated not only for individual trips, but for trip chains, understood as the total distance an
individual travels on all the trips that start and end at the same place [
12
]. The threshold for a trip chain
in this study is calculated as the 80th percentile of the distribution of the sum of distances travelled by
an individual using active transport modes on the survey day, as proposed by [11].
Two additional criteria are considered as limiting factors when allocating a car trip as being
transferable to cycling. The first is that the person knows how to ride a bike and has one at home,
and the second is that the individual travelling on the short car trip must have travelled by bicycle,
for the same purpose, in the month before the day of the survey. All trips that do not meet these
limiting criteria are discarded from the car trips that are potentially replaceable by bicycle trips.
The same thresholds are also calculated using the travel time declared by the survey respondents
as a data source for travel distance, in order to provide supplementary information on the
perceived times.
2.2.2. Analysing the Barriers to Cycling
After establishing the criteria to define the SCT that are replaceable by active modes, the next
step was to identify the individuals in the HTS who used the car for this type of trip, and assess the
perceived barriers to bicycle use in this subset of the sample. This part of the study is limited to the
cycling active mode, due to the availability of the data in the HTS [24].
The subset of people who travelled in SCT is divided into two groups corresponding to habitual
and non-habitual bicycle users. The perceived barriers to bicycle use as a means of transportation
are analysed through the responses to a set of questions included in the HTS [
24
] for this purpose.
These questions are shown in Table 1. It should be noted that there are other factors that could be
considered as barriers to bicycle use, such as the lack of adequate maps and routing information,
or the absence of cycle retailers, among others [
28
]. However, in order to constrain the extension of the
survey, only ten barriers were considered.
Table 1. Questions for analysing perceived barriers to bicycle use as a mean of transportation.
Code Question Answer Type
How far do/would the following facts limit your regular trips by bicycle?
B1 Lack of cycle lane 7-point Likert scale a
B2 Long distances 7-point Likert scale a
B3 Lack of safe parking 7-point Likert scale a
B4 Slope 7-point Likert scale a
B5 Riding a bicycle in traffic 7-point Likert scale a
B6 Lack of showers or lockers 7-point Likert scale a
B7 Safety when manoeuvring 7-point Likert scale a
B8 Your physical condition 7-point Likert scale a
B9 Repairing a puncture 7-point Likert scale a
B10 Helmet use 7-point Likert scale a
Rate the following statements:
A1 In the next six months I will increase the use of the bicycle on my habitual trips 7-point Likert scale b
A2 Travelling by bicycle in Vitoria-Gasteiz is efficient, convenient and safe 7-point Likert scale b
a
7-point Likert scale from 1 (not limiting) to 7 (very limiting);
b
7-point Likert scale from 1 (totally disagree) to 7
(totally agree).

Sustainability 2018,10, 3510 5 of 13
Descriptive statistics are then applied to analyse each group’s answers to the questions on the
perceived barriers (mean, median, standard deviation and frequency). Contingency tables are used
to determine whether the perception of each barrier is homogeneous among the different groups.
Specifically, 12 tables are built to relate the data on two categorical variables by comparing the observed
frequencies (in this study, each table relates the three population groups described above, with each
of the perceived barriers B1–B10 and the statements A1 and A2 in Table 1). To perform the analyses,
the variables relating to the perceived barriers are recorded in three categories; not very limiting (1, 2, 3),
neutral (4), and very limiting (5, 6, 7). The relationship between the variables in the contingency tables
is measured with Cramér’s V statistic. This is a correction of the Chi square that takes into account
the sample size. The Chi square contrast compares the observed frequencies with the expected
frequencies [
29
], and indicates any relationship between the variables. Cramér’s V [
30
] adopts values
of between 0 and 1, and indicates not only if there is a relationship between the variables, but also its
degree (0 for non-dependent variables, and 1 for the maximum relationship between the variables).
V=sχ2
N(k−1). (1)
where:
χ2is the Pearson chi-square statistic
Nis the total number of observations
kis the smaller of the number of categories of either variables
3. Results
3.1. Distance and Time Thresholds for the Potential Replacement of SCT by Active Modes
Tables 2and 3show the thresholds that define the SCTs that could be replaced by active modes
for the whole sample and for the different population groups. Table 2distinguishes between the values
obtained for individual walking trips and for walking trips in a chain. The tables also include the trip
frequency (n) and average travel times obtained through the Google Maps API [
27
], and the average
time perceived by the respondents in the HTS sample. Table 2shows that the average times for walking
trips calculated using the Google Maps API are lower than the perceived times, leading to a greater
difference in the older population groups (Table 3). Distances of between 1.6 and 2 km are observed
in the thresholds for individual trips, with little difference between age and sex groups, while the
definition threshold for a trip chain is between 5 and 8 km.
Table 2. Thresholds for individual walking trips and walking trips as part of a chain.
Age Sex Freq. Walking
Trips [n]
Average API
Time [min]
Average Perceived
Time [min]
Threshold (80th
Percentile) [km]
Chain Trip
Threshold [km]
All (18–84) M 2435 13.30 17.38 1.81 6.38
F 4835 12.36 16.57 1.62 5.93
18–34 M 425 13.76 14.21 1.65 5.26
F 686 13.76 15.85 1.73 5.07
35–49 M 657 12.81 14.97 1.66 5.38
F 1708 12.45 15.83 1.60 6.09
50–64 M 586 13.61 18.27 1.83 6.60
F 1400 12.61 16.61 1.68 6.43
65–84 M 767 13.16 20.47 2.04 7.86
F 1041 10.95 18.23 1.60 6.00

Sustainability 2018,10, 3510 6 of 13
Table 3. Cycling trip thresholds.
Age Sex Freq. Cycling
Trips [n]
Average Times
(API) [min]
Perceived
Time [min]
Threshold (80th
Percentile) [km]
All (18–84) M 725 9.65 16.44 3.70
F 618 8.61 16.47 3.23
18–34 M 338 9.63 16.50 3.56
F 275 8.81 16.65 3.45
35–49 M 258 10.7 16.51 4.41
F 263 8.91 16.07 3.28
50–64 M 97 7.51 15.30 3.53
F 58 6.75 15.29 2.71
65–84 M 32 7.43 18.50 3.58
F 22 6.51 22 3.56
In the case of cycling (Table 3), trip chains that could be made by bicycle were not considered,
since the threshold could not be seen as a limiting factor. This is because there were very few car
trip chains over the distance threshold defined as the 80th percentile of the distribution of distances
travelled in bicycle trip chains. However, for individual trips, the results show that the threshold
for a short car trip which could be replaced by cycling is between 3 and 4.5 km, depending on the
population group.
The application of these thresholds to car trips in the HTS sample produces a percentage of
between 14% and 29% of car trips that can potentially be walked, with a higher percentage generally
seen among women (Table 4). The results in Table 4show that 50% to 70% of the SCT are below the
cycling distance thresholds in Table 3. However, once the limiting factors are considered (bicycle
availability and use in the previous month), the number of SCTs that are transferable to the bicycle
decreases significantly (see SCTs transferable to cycling in Table 4). This reduction is higher for women
(from 59.93% to 19.21% in all age groups) than for men (62.25% to 31.47% in all age groups) and high
age groups (e.g., from 64.15% to 16.37% in 50-64 male age group). The last column in Table 4shows
the total number of trips that could be replaced by walking or cycling trips according to the criteria
established in Section 2.2.2.
Table 4. SCTs transferable to active modes.
Age Sex Car Trips
(Total) [N]
SCTs
Transferable to
Walking [N (%)]
Short Car Trips
below the Cycling
Threshold [N (%)]
Short Car Trips
Transferable to
Cycling [N (%)]
Short Car Trips
Transferable to Active
Modes (Total) [N (%)]
All (18–84) M 1436 293 (20.40) 894 (62.25) 452 (31.47) 602 (41.92)
F 1837 380 (20.68) 1101 (59.93) 353 (19.21) 621 (33.80)
18–34 M 363 50 (13.77) 224 (61.70) 150 (41.32) 161 (44.35)
F 356 63 (17.69) 212 (59.55) 101 (28.37) 134 (37.64)
35–49 M 643 100 (15.55) 445 (69.20) 252 (39.19) 294 (45.72)
F 1036 216 (20.84) 637 (61.48) 216 (20.84) 372 (35.90)
50–64 M 226 45 (19.91) 145 (64.15) 37 (16.37) 86 (38.05)
F 308 76 (24.67) 162 (52.59) 28 (9.09) 89 (28.89)
65–84 M 161 46 (28.57) 92 (57.14) 17 (10.55) 51 (31.67)
F 96 25 (26.04) 70 (72.91) 0 (0) 25 (26.04)
3.2. Perceived Barriers to Cycling
After analysing the SCTs transferable to active modes, a study was made of the perceived barriers
to the use of the bicycle by the different groups in Section 2.2. The descriptive statistics shown in
Table 5and Figures 2and 3reveal that the respondents who had not used a bicycle in the last month
perceived a greater limitation to barriers than those who had used one and had made at least a short
car trip the day before the survey.

Sustainability 2018,10, 3510 7 of 13
Table 5. Contingency tables and descriptive statistics of perceived barriers.
Reclassification of Barriers
Barrier User Groups Mean
(±SD) Median Not very Limiting
[n (%)]
Neutral
[n (%)]
Very Limiting
[n (%)] Cramér ’s V
B1 SCT user c5.14 (±1.87) 6 58 (18.01) 45 (13.97) 219 (68.01) 0.11 *
SCT/Bicycle user d4.7 (±1.89) 5 40 (24.39) 26 (15.85) 98 (59.75)
Bicycle user e4.39 (±1.83) 5 219 (29.12) 151 (20.07) 382 (50.79)
B2 SCT user c5.16 (±1.67) 6 50 (15.52) 43 (13.35) 229 (71.11) 0.22 *
SCT & Bicycle user d4.87 (±1.74) 5 32 (19.51) 31 (18.9) 101 (61.58)
Bicycle user e3.83 (±1.81) 4 308 (40.95) 164 (21.8) 280 (37.23)
B3 SCT user c5.32 (±1.68) 6 49 (15.21) 40 (12.42) 233 (72.36) 0.14 *
SCT & Bicycle user d4.99 (±1.77) 5 30 (18.29) 28 (17.07) 106 (64.63)
Bicycle user e4.45 (±1.88) 5 232 (30.85) 132 (17.55) 388 (51.59)
B4 SCT user c5.08 (±1.68) 5 52 (16.14) 62 (19.25) 208 (64.59) 0.25 *
SCT & Bicycle user d4.37 (±1.79) 4 49 (29.87) 34 (20.73) 81 (49.39)
Bicycle user e3.45 (±1.72) 4 375 (49.86) 173 (23) 204 (27.12)
B5 SCT user c6.17 (±1.38) 7 22 (6.83) 20 (6.21) 280 (86.95) 0.19 *
SCT & Bicycle user d5.89 (±1.51) 6 13 (7.92) 16 (9.75) 135 (82.31)
Bicycle user e4.88 (±1.8) 5 170 (22.6) 121 (16.09) 461 (61.3)
B6 SCT user c4.49 (±1.9) 5 93 (28.88) 65 (20.18) 164 (50.93) 0.18 *
SCT & Bicycle user d3.73 (±1.88) 4 71 (43.29) 32 (19.51) 61 (37.19)
Bicycle user e3.19 (±1.85) 3 433 (57.57) 126 (16.75) 193 (25.66)
B7 SCT user c4.92 (±1.78) 5 55 (17.08) 63 (19.56) 204 (63.35) 0.19 *
SCT & Bicycle user d4.65 (±1.68) 5 31 (18.9) 38 (23.17) 95 (57.92)
Bicycle user e3.88 (±1.78) 4 292 (38.82) 184 (24.46) 276 (36.7)
B8 SCT user c4.47 (±1.9) 5 93 (28.88) 62 (19.25) 167 (51.86) 0.24 *
SCT & Bicycle user d3.74 (±1.99) 4 74 (45.12) 26 (15.85) 64 (39.02)
Bicycle user e2.86 (±1.84) 2 504 (67.02) 84 (11.17) 164 (21.8)
B9 SCT user c5.71 (±1.77) 7 41 (12.73) 36 (11.18) 245 (76.08) 0.24 *
SCT & Bicycle user d5.17 (±1.88) 6 31 (18.9) 22 (13.41) 111 (67.68)
Bicycle user e3.91 (±2.04) 4 322 (42.81) 126 (16.75) 304 (40.42)
B10 SCT user c3.34 (±2.14) 3 177 (54.96) 48 (14.9) 97 (30.12) 0.05
SCT & Bicycle user d3.45 (±2.18) 3.5 82 (50) 27 (16.46) 55 (33.53)
Bicycle user e3.57 (±2.13) 4 357 (47.47) 139 (18.48) 256 (34.04)
* Statistically significant differences among groups (0.01);
c
SCT user: respondent who travelled in a SCT transferable
to active modes and has not ridden a bicycle in the last month
d
SCT user/Bicycle user: respondent who travelled
in a SCT transferable to active modes and has ridden a bicycle in the last month
e
Bicycle user: respondent who
cycles on their habitual trips.
Sustainability 2018, 10, x FOR PEER REVIEW 8 of 13
user: respondent who travelled in a SCT transferable to active modes and has ridden a bicycle in the
last month e Bicycle user: respondent who cycles on their habitual trips.
Table 6. Contingency tables and descriptive statistics of limiting factors.
Reclassification of Barriers
Barrier User Groups Mean
(±SD) Median Totally Disagree
[n (%)]
Neutral
[n (%)]
Totally Agree
[n (%)] Cramér’s V
A1 SCT user
c 2.31 (±1.95) 1 243 (75.46) 24 (7.45) 55 (17.08) 0.30 *
SCT & Bicycle user d 4.37 (±2.21) 5 50 (30.48) 25 (15.24) 89 (54.26)
Bicycle user e 4.67 (±2.26) 5 205 (27.26) 97 (12.89) 450 (59.84)
A2 SCT user
c 4.57 (±1.76) 5 68 (21.11) 73 (22.67) 181 (56.21) 0.14 *
SCT & Bicycle user d 4.95 (±1.45) 5 20 (12.19) 35 (21.34) 109 (66.46)
Bicycle user e 5.36 (±1.35) 5.5 60 (7.97) 129 (17.15) 56 (34.86)
* Statistically significant differences among groups (0.01); c SCT user: respondent who travelled in a
SCT transferable to active modes and has not ridden a bicycle in the last month; d SCT user/Bicycle
user: respondent who travelled in a SCT transferable to active modes and has ridden a bicycle in the
last month; e Bicycle user: respondent who cycles on their habitual trips.
Figure 2. mean, median, and standard deviation of perceived barriers.
Figure 3. relative frequency of barrier perception.
0
1
2
3
4
5
6
7
8
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 A1 A2
Mean ± SD Median
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 A1 A2
Not very limiting Neutral Very limiting
Figure 2. Mean, median, and standard deviation of perceived barriers.

Sustainability 2018,10, 3510 8 of 13
Sustainability 2018, 10, x FOR PEER REVIEW 8 of 13
user: respondent who travelled in a SCT transferable to active modes and has ridden a bicycle in the
last month e Bicycle user: respondent who cycles on their habitual trips.
Table 6. Contingency tables and descriptive statistics of limiting factors.
Reclassification of Barriers
Barrier User Groups Mean
(±SD) Median Totally Disagree
[n (%)]
Neutral
[n (%)]
Totally Agree
[n (%)] Cramér’s V
A1 SCT user
c 2.31 (±1.95) 1 243 (75.46) 24 (7.45) 55 (17.08) 0.30 *
SCT & Bicycle user d 4.37 (±2.21) 5 50 (30.48) 25 (15.24) 89 (54.26)
Bicycle user e 4.67 (±2.26) 5 205 (27.26) 97 (12.89) 450 (59.84)
A2 SCT user
c 4.57 (±1.76) 5 68 (21.11) 73 (22.67) 181 (56.21) 0.14 *
SCT & Bicycle user d 4.95 (±1.45) 5 20 (12.19) 35 (21.34) 109 (66.46)
Bicycle user e 5.36 (±1.35) 5.5 60 (7.97) 129 (17.15) 56 (34.86)
* Statistically significant differences among groups (0.01); c SCT user: respondent who travelled in a
SCT transferable to active modes and has not ridden a bicycle in the last month; d SCT user/Bicycle
user: respondent who travelled in a SCT transferable to active modes and has ridden a bicycle in the
last month; e Bicycle user: respondent who cycles on their habitual trips.
Figure 2. mean, median, and standard deviation of perceived barriers.
Figure 3. relative frequency of barrier perception.
0
1
2
3
4
5
6
7
8
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 A1 A2
Mean ± SD Median
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
SCT user
Bicycle user
SCT / Bicycle user
B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 A1 A2
Not very limiting Neutral Very limiting
Figure 3. Relative frequency of barrier perception.
The frequencies of the reclassified barrier values are shown in the 12 contingency tables in Tables 5
and 6and in Figure 3. It was found that 86.95%, 82.31% and 61.3% of the user groups, respectively,
perceive riding a bicycle in traffic as the most limiting barriers. Repairing a puncture (B9), the lack of
parking at destination (B3), long distances (B2), and the lack of cycle lanes (B1) are also perceived as
important barriers by three user groups. Regarding the value of the Cramér’s V statistic, which gives
the degree of relationship of the variables, the difference between the groups is significant for all the
barriers except for question B10 (use of helmet). Four questions show a closer relationship with the
type of user group: questions B2, B4, B8, and B9 (related to distance travelled, slope, physical condition
and punctures, respectively). In all cases, the percentage of people who consider a barrier as very
limiting is higher for car users than for those who do not use a bicycle.
Table 6. Contingency tables and descriptive statistics of limiting factors.
Reclassification of Barriers
Barrier User Groups Mean (±SD) Median Totally Disagree
[n (%)]
Neutral
[n (%)]
Totally Agree
[n (%)] Cramér’s V
A1 SCT user c2.31 (±1.95) 1 243 (75.46) 24 (7.45) 55 (17.08) 0.30 *
SCT & Bicycle user d4.37 (±2.21) 5 50 (30.48) 25 (15.24) 89 (54.26)
Bicycle user e4.67 (±2.26) 5 205 (27.26) 97 (12.89) 450 (59.84)
A2 SCT user c4.57 (±1.76) 5 68 (21.11) 73 (22.67) 181 (56.21) 0.14 *
SCT & Bicycle user d4.95 (±1.45) 5 20 (12.19) 35 (21.34) 109 (66.46)
Bicycle user e5.36 (±1.35) 5.5 60 (7.97) 129 (17.15) 56 (34.86)
* Statistically significant differences among groups (0.01);
c
SCT user: respondent who travelled in a SCT transferable
to active modes and has not ridden a bicycle in the last month;
d
SCT user/Bicycle user: respondent who travelled
in a SCT transferable to active modes and has ridden a bicycle in the last month;
e
Bicycle user: respondent who
cycles on their habitual trips.
The question related to intention to increase bicycle use (A1) is also closely related with the type
of group. It was found that 75.46% of these car users totally disagreed with increasing the use of a
bicycle on their habitual trips. In contrast, 59.84% of bicycle users had the intention to increase the use
of the bicycle on their habitual trips. Most SCT users, SCT & bicycle users, and bicycle users think that
it as an efficient, convenient, and safe experience (56.21%, 66.46%, and 34.86% respectively). It should
be noted that the percentage of people with this opinion is lower in bicycle users than in SCT users.

Sustainability 2018,10, 3510 9 of 13
4. Discussion
Based on the results of a mobility survey [
24
], this study analysed the SCTs that could be replaced
by active modes, and the main perceived barriers to the use of bicycles by different population groups.
It should be noted that the thresholds obtained to define a SCT that could be replaced by a walking
trip are within the ranges defined by previous studies [
6
,
11
]. However, the thresholds obtained for
the definition of travel chains that can be done on foot are considerably higher than those calculated
with a similar methodology by [
11
]. This may be due to the different typology of the case studies
analysed, since the present work focuses on a medium-sized city in which more than half the total
trips are done on foot. The segmentation of the sample used to calculate the thresholds has been
limited to age and gender; further works may use other personal and socioeconomic variables such as
socioeconomic status, education or level occupation, which have been demonstrated to be relevant in
active mobility [31,32].
The average perceived distances are higher than the ones calculated by the Google Maps API
for all age groups, and this difference is greater as the age of the group increases (Tables 2and 3).
One possible explanation is that the API uses a single speed to calculate travel times, whereas in reality,
different age groups may walk at different average speeds [
33
]. The difference between the perceived
travel time and the travel time calculated by the API could be caused by the fact that some trips may
include stops or alternative routes to the optimal route calculated by the API. The greatest difference
observed in the case of the bicycle may be due to the parking time at origin and at destination.
Once the thresholds were determined, they were used to screen the short car trips to determine
which ones could potentially be substituted by active modes. It can be seen that when only bicycle
distance thresholds are taken into account, a large percentage of trips can be considered as replaceable
by active modes. However, when the limiting factors are included for these trips (availability and bike
use), these percentages decrease significantly, although they are still important. It should be noted
that the limiting factors (bicycle availability and use in the previous month) decrease significantly the
car trips transferable to cycling for seniors, and it is more marked for women than for men in all age
groups. Similarly, the sex variable also appears to have an influence when the limiting factors are
included, with men making a greater proportion of short trips that could be substituted by cycling.
This may be due to the fact that gender and age variables influence the use of the bicycle, as was
already identified by [
16
], and also seen in the results in Table 3. After including the limiting factors
for trips that could be made by bicycle, the remaining trips are added to the trips that could be made
on foot. According to our results, between 30% and 40% of car trips made in Vitoria-Gasteiz could be
replaced by active modes.
The next step was to investigate why these trips—which, according to our results, could be made
by active modes—were made by car. A study was made of the perceived barriers for the use of the
bicycle for the three categories described in the methodology. We were unable to analyse the perceived
barriers for journeys that were potentially transferable to walking because the HTS survey did not ask
about the perceived barriers for this mode of transport. The results of the statistical analyses in Tables 5
and 6show significant differences between people who regularly use the bicycle and car for short trips,
and, of these, also between bicycle and non-bicycle users. The differences in the perception of barriers
are significant in all cases, except for the use of a helmet (B10), for which the three groups find the
barrier less limiting (see Figure 3). It should be noted that helmet use is the only barrier considered
in this study that is perceived as more limiting by habitual bicycle users, possibly owing to a more
positive perception of personal safety when using the bicycle, and to the fact that these users have
experienced first-hand the inconvenience of having to carry around the helmet after a trip.
The barriers perceived as the most limiting by car users who do not use the bicycle concern
personal safety, such as riding a bicycle in traffic (B5) and the lack of cycle lanes (B1). Practical issues
such as repairing a puncture (B9), the lack of parking at destination (B3), and long distances (B2) are
also perceived as important barriers by car users. Car users who also use the bicycle, albeit occasionally,
show a similar pattern, but generally perceive most of the barriers as less limiting. However, the barriers

Sustainability 2018,10, 3510 10 of 13
most closely related to physical condition (B4 and B8) and technique (B9) reveal the greatest dependence
between the perceived limitation and the user group. As expected, the group that perceives the barriers
associated with cycling as least limiting (with the exception of B10) is the regular bicycle users.
One significant finding is that the greatest difference between groups in terms of the statement on
a future increase in bicycle use (A1) is between car users, who have never used a bicycle, and the other
two groups. It is interesting that car users who also use a bicycle as a means of transport show a clear
intention to increase their bicycle use in the coming months, since it may indicate a decrease in the use
of the car for these short trips in the near future, thus pinpointing a target group for intervention. It is
also worth noting that although car users who do not use a bicycle show no clear intention of adopting
this habit, they still perceive the use of the bicycle as an efficient, convenient, and safe experience,
suggesting that, over time, these users could be encouraged to take up cycling on their regular trips.
These results should encourage policy-makers to continue with the current strategies contemplated
in the Mobility and Public Space Plan and the Cycling Mobility Transport Plan of Vitoria-Gasteiz.
However, our results suggest that there is still a great margin for improvement. The great percentage
of SCTs that are transferable to active modes indicate that it could be worth concentrating policy
efforts on reducing them. Other studies have suggested that, in some cases, the increase of cycling
modal share could come mainly from the substitution of public transport trips or new trips [
15
].
However, out results in Vitoria-Gasteiz confirm that cycling modal share could also be increased from
the substitution of SCTs. In this respect, a possible future research line could be to incorporate the
effects of policies aimed at integrating active modes and public transport—such as provision of bike
racks on buses, accommodation of bikes on rail vehicles, and bike parking [
34
]—in the transference of
trips towards cycling and walking.
We have defined the potential reduction in car use in Vitoria-Gasteiz, but we have not addressed
the evaluation of the health benefits obtained from a shift from car use to cycling and walking,
which includes the change in exposure to ambient air pollution for the individuals who change their
transportation mode, their health benefit, the health benefit for the general population due to reduced
pollution, and the influence of accident risk [
35
]. Further research is needed to examine these effects in
the case study.
This study has at least two more limitations that could be addressed in further research. The first
is that we did not have access to information on the perceived barriers to walking. Although this
means of transport has less potential for replacing short trips, it is an inclusive means of transport
that can be used by almost anyone. We therefore consider it important in future research to analyse
perceptions of pedestrian barriers, and relate them to the potential substitution of short car trips.
The second limitation is that the distances used to define the SCT thresholds were calculated from
minimum-cost routes obtained through the Google Maps API algorithm. In addition to the minimum
route distance, pedestrian and cycling routes could be also influenced by other factors in the built
environment. A possible consideration in future research would be to use real routes gathered from
GPS surveys, or routes derived from the analysis of the extensive information available from future 5G
mobile networks.
5. Conclusions
This study answers the research questions formulated. We have analysed the modal shift to active
transport modes, and specifically identified the car trips that could potentially be transferred to active
modes such as walking and cycling. Finally, we have studied the barriers perceived by users of SCT
that hinder this shift in transport mode.
Based on a mobility survey with pedestrian origin and destinations, and distances calculated
using the Google Maps API, a distance threshold approach was applied in Vitoria-Gasteiz in order to
identify car trips that could potentially be replaced by walking and cycling trips. Using descriptive
statistics, the answers of habitual and non-habitual bicycle users to a set of questions in the HTS
enabled us to identify the perceived barriers for the use of a bicycle.

Sustainability 2018,10, 3510 11 of 13
The results suggest a high potential for reducing car use on these types of trips. The threshold
for a short car trip which could be replaced by walking is between 1.6 and 2 km, with little difference
between age and sex groups. In the case of cycling, the threshold is between 3 and 4.5 km. The results
show that between 30% and 40% of car trips made in Vitoria-Gasteiz could be replaced by active modes.
Specifically, between 14% and 29% of car trips can potentially be walked, with a higher percentage
among women, and between 50 and 70% could be cycled. In this second case, the limiting factors
considered (bicycle availability and use in the previous month) significantly decrease the number of
SCTs that are transferable, depending of the age and sex. This reduction is high for women and high
age groups.
These findings highlight significant differences between regular bicycle and car users for short
trips. The barriers perceived as the most limiting by car users concern personal safety, such as riding a
bicycle in traffic, and the lack of cycle lanes. It was found that 71% and 56%, respectively, of the total
users consider them very limiting. Practical issues (repairing, parking and distance) are also perceived
as important barriers by car users. For all barriers except helmet use, the percentage of people who
consider a barrier as very limiting is higher for car users than for those who do not use a bicycle.
Although they see it as an efficient, convenient, and safe experience (56% are totally agree),
they show no clear intention of adopting this mode. Only 17.08% of car users totally agreed with
increasing the use of the bicycle on their habitual trips. In contrast, 59.84% of bicycle users had
the intention to increase the use of the bicycle on their habitual trips. It should be noted also that
the percentage of people that think that cycling is efficient, convenient, and safe is lower among
bicycle users (34.86%) than in SCT users (56.21%). In our opinion, all these results provide valuable
information for implementing measures to promote the replacement of SCT by cycling.
Author Contributions:
This paper was conceived and written jointly by the authors. J.D. formulated the research
questions, performed the calculations, analysed the data, and wrote the text; B.M. completed the methods,
formulated the research questions, and wrote the text. E.O. was responsible for the research motivations,
the structure, extracting conclusions and for the comprehensiveness of the paper.
Funding:
This paper was produced under the framework of the following projects: “Desarrollo de Aplicaciones
SIG para la Implementación de Indicadores de Fragmentación Urbana y Mejora de la Movilidad” funded by the
Universidad Politécnica de Madrid, research project no. RP151320028, and “DESPACIO” (TRA2017-88058-R)
funded by the Spanish Ministry of Science in the “Programa Estatal de I+D+i Orientada a los Retos de la Sociedad”.
The authors wish to acknowledge the financial support of the Department of Education of the Madrid Region
and the European Social Fund (PEJD-2016/AMB-2671), and the Spanish Ministry of Education (FPU16/01557),
which awarded grants to one of the authors (Javier Delso).
Acknowledgments:
The authors would like to thank the Centro de Estudios Ambientales (CEA Victoria-Gasteiz)
and the “TRANSBICI” Project (TRA2010-17035) for supplying the HTS, and María Gómez-Elvira for her help
during this research.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Rueda, S. La Supermanzana, Nueva Célula Urbana para la Construcción de un Nuevo Modelo Funcional y
Urbanístico de Barcelona; BCN Ecologia: Barcelona, Spain, 2006; Available online: http://bcnecologia.net/
sites/default/files/proyectos/la_supermanzana_nueva_celula_poblenou_salvador_rueda.pdf (accessed
on 15 March 2018).
2.
European Environmental Agency. Transitions towards a More Sustainable Mobility System. TERM 2016:
Transport Indicators Tracking Progress towards Environmental Targets in Europe. European Environmental
Agency Report. 2016. No. 34/2016. Available online: https://www.eea.europa.eu//publications/term-
report-2016 (accessed on 10 January 2018).
3.
European Commission. Links between Noise and Air Pollution. Science for Environment Policy. In-depth
Report 13. 2016. Available online: http://ec.europa.eu/environment/integration/research/newsalert/pdf/
air_noise_pollution_socioeconomic_status_links_IR13_en.pdf (accessed on 18 February2018).

Sustainability 2018,10, 3510 12 of 13
4.
Mueller, N.; Rojas-Rueda, D.; Basagaña, X.; Cirach, M.; Cole-Hunter, T.; Dadvand, P.; Donaire-Gonzalez, D.;
Foraster, M.; Gascon, M.; Martinez, D.; et al. Urban and Transport Planning Related Exposures and Mortality:
A Health Impact Assessment for Cities. Environ. Health Perspect. 2017,125, 89–96. [CrossRef] [PubMed]
5. Sugiyama, T.; Merom, D.; Van der Ploeg, H.P.; Corpuz, G.; Bauman, A.; Owen, N. Prolonged sitting in cars:
Prevalence, socio-demographic variations, and trends. Prev. Med. 2012,55, 315–318. [CrossRef] [PubMed]
6.
Morency, C.; Demers, M.; Poliquin, E. Shifting short motorized trips to walking: The potential of active
transportation for physical activity in Montreal. J. Transp. Health 2014,1, 100–107. [CrossRef]
7.
Beckx, C.; Broekx, S.; Degraeuwe, B.; Beusen, B.; Int Panis, L. Limits to active transport substitution of SCT.
Transp. Res. D. Transp. Environ. 2013,22, 10–13. [CrossRef]
8.
André, J.-M.; Joumard, R. Modelling of Cold Start Excess Emissions for Passenger Cars. INRETS Report
LTE. 2005. Available online: https://hal.archives-ouvertes.fr/hal-00917071/document (accessed on 15
March 2018).
9.
Favez, J.Y.; Weilenmann, M.; Stilli, J. Cold start extra emissions as a function of engine stop time: Evolution
over the last 10 years. Atmos. Environ. 2009,43, 996–1007. [CrossRef]
10.
Hanson, S. Gender and mobility: New approaches for informing sustainability. Gend. Place Cult.
2010
,17,
5–23. [CrossRef]
11.
Cole, R.; Turrell, G.; Koohsari, M.J.; Owen, N.; Sugiyama, T. Prevalence and correlates of walkable short car
trips: A cross-sectional multilevel analysis. J. Transp. Health 2017,4, 73–80. [CrossRef]
12.
Ye, X.; Pendyala, R.M.; Gottardi, G. An exploration of the relationship between mode choice and complexity
of trip chaining patterns. Transp. Res. B. Meth. 2007,41, 96–113. [CrossRef]
13.
Badland, H.; Schofield, G. Perceptions of replacing car journeys with non-motorized travel: Exploring
relationships in a cross-sectional adult population sample. Prev. Med.
2006
,43, 222–225. [CrossRef]
[PubMed]
14.
Geus, B.; De Bourdeaudhuij, I.; Jannes, C.; Meeusen, R. Psychosocial and environmental factors associated
with cycling for transport among a working population. Health Educ. Res.
2008
,23, 697–708. [CrossRef]
[PubMed]
15.
Handy, S.; Van Wee, B.; Kroesen, M. Promoting cycling for transport: research needs and challenges. Transp.
Rev. 2014,34, 4–24. [CrossRef]
16.
Barberán, A.; Monzón, A. How did bicycle share increase in Vitoria-Gasteiz? In Proceedings of the XII
Conference on Transport Engineering, CIT, Valencia, Spain, 7–9 June 2016.
17.
Bauman, A.; Rissel, C.; Garrard, J.; Ker, I.; Speidel, R.; Fishman, E. Cycling: Getting Australia Moving—Barriers,
Facilitators and Interventions to Get More Australians Physically Active through Cycling; 31st Australasian
Transport Research Forum; Cycling Promotion Fund: Melbourne, Australia, 2008; pp. 593–601.
18.
Fraser, S.D.S.; Lock, K. Cycling for transport and public health: A systematic review of the effect of the
environment on cycling. Eur. J. Public Health 2011,21, 738–743. [CrossRef] [PubMed]
19.
Parkin, J.; Wardman, M.; Page, M. Models of perceived cycling risk and route acceptability. Accid. Anal. Prev.
2007,39, 364–371. [CrossRef] [PubMed]
20.
Michelle, D.; Chris, R.; Beverley, L.L. All Dressed Up and Nowhere to Go?: A Qualitative Research Study of
the Barriers and Enablers to Cycling in Inner Sydney. Road Transp. Res. J. Aust. N. Z. Res. Pract.
2007
,16,
42–52.
21.
Piatkowski, D.P.; Marshall, W.E. Not all prospective bicyclists are created equal: The role of attitudes,
socio-demographics, and the built environment in bicycle commuting. Travel Behav. Soc.
2015
,2, 166–173.
[CrossRef]
22.
Mandic, S.; Hopkins, D.; García Bengoechea, E.; Flaherty, C.; Williams, J.; Sloane, L.; Spence, J.C. Adolescents’
perceptions of cycling versus walking to school: Understanding the New Zealand context. J. Transp. Health
2017,4, 294–304. [CrossRef]
23.
Muñoz, B.; Rondinella, G. Informe de Evaluación del Plan de Movilidad Sostenible y Espacio Público y del Plan
Director de Movilidad Ciclista de Vitoria-Gasteiz; Centro de Estudios Ambientales: Vitoria-Gasteiz, Spain, 2017;
Available online: http://www.vitoria-gasteiz.org/wb021/http/contenidosEstaticos/especial/2017/0406/
PMSEP_evaluacion_2017.pdf (accessed on 3 September 2018). (In Spanish)

Sustainability 2018,10, 3510 13 of 13
24.
Vitoria-Gasteiz. Análisis de los resultados de la encuesta de movilidad 2014 de Vitoria-Gasteiz realizada
por TRANSyT (Universidad Politécnica de Madrid) para el Centro de Estudios Ambientales (CEA). 2014.
Available online: https://blogs.vitoria- gasteiz.org/medios/files/2015/01/an%25C3%25A1lisis-encuesta-
movilidad-2014.pdf (accessed on 10 December 2017).
25.
Rondinella, G. Considering Cycling for Commuting: The Role of Mode Familiarity: An Exploration on the
(Circular) Relation between Cycling Behaviours and Attitudes toward Cycling in Vitoria-Gasteiz. Ph.D.
Dissertation, Universidad Politécnica de Madrid, Madrid, Spain, July 2015.
26.
Muñoz López, B. Integrating Bicycle Option in Mode Choice Models through Latent Variables. Ph.D.
Dissertation, Universidad Politécnica de Madrid, Madrid, Spain, July 2016.
27. Google Maps API. Terms of Use. 2018. Available online: http://www.google.com/apis/maps/terms.html
(accessed on 15 June 2018).
28.
Scheepers, C.E.; Wendel-Vos, G.C.W.; den Broeder, J.M.; van Kempen, E.E.M.M.; van Wesemael, P.J.V.;
Schuit, A.J. Shifting from car to active transport: A systematic review of the effectiveness of interventions.
Transp. Res. Part A 2014,70, 264–280. [CrossRef]
29. Peña, D. Fundamentos de Estadistica; Alianza Editorial: Madrid, Spain, 2010; ISBN 978-84-206-8380-5.
30. Cramér, H. Mathematical Methods of Statistics; Princeton University Press: Princeton, NJ, USA, 1946.
31.
Goodman, A. Walking, Cycling and Driving to Work in the English and Welsh 2011 Census: Trends,
Socio-Economic Patterning and Relevance to Travel Behaviour in General. PLoS ONE
2013
,8, e71790.
[CrossRef] [PubMed]
32.
Simons, D.; De Bourdeaudhuij, I.; Clarys, P.; de Geus, B.; Vandelanotte, C.; Van Cauwenberg, J.; Deforche, B.
Choice of Transport Mode in Emerging Adulthood: Differences Between Secondary School Students,
Studying Young Adults and Working Young Adults and Relations with Gender, Ses and Living Environment.
Transp. Res. Part A Policy Pract. 2017,103, 172–184. [CrossRef]
33.
Gates, T.; Noyce, D.; Bill, A.; Van Ee, N. Recommended walking speeds for pedestrian clearance timing
based on pedestrian characteristics. Transp. Res. Rec. 2006,1982, 38–47. [CrossRef]
34.
Pucher, J.; Buehler, R. Integrating bicycling and public transport in North America. J. Public Transp.
2009
,12,
79–104. [CrossRef]
35.
Rabl, A.; De Nazelle, A. Benefits of shift from car to active transport. Transp. Policy
2012
,19, 121–131.
[CrossRef]
©
2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).