Content uploaded by Sameer Abu-Eisheh
Author content
All content in this area was uploaded by Sameer Abu-Eisheh on Feb 29, 2024
Content may be subject to copyright.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
2590-1982/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Trip generation model for a developing city in an emerging country
Sameer Abu–Eisheh
a
, Mohammad S. Ghanim
b
,
*
, Alaa Dodeen
a
a
An-Najah National University, PO BOX 7, Nablus, Palestine
b
Ministry of Transport, PO BOX 24455, Doha, Qatar
ARTICLE INFO
Keywords:
Trip generation
Transportation planning
Multivariable regression
Developing cities
Land use
Palestine
ABSTRACT
This paper presents the development of trip generation models in an emerging country. In this paper, trip
generation models for a medium-size Palestinian city, the City of Jericho, are established. This study is the rst
comprehensive city-wide trip generation study that was done within the Palestinian territories. Moreover, the
study considers mixed land use while developing trip generation models. Household demographic, socioeco-
nomic, land use and trip data were collected to generate models to represent trip generation for trips with
different purposes, such as number of daily trips for educational, shopping, social, and recreational purposes.
Stratied and random household samples are drawn from the study area, covering 14 different trafc analysis
zones. All the households were visited, and personal face-to-face interviews were conducted to collect de-
mographic, socioeconomic and travel behavior information. A multivariable regression analysis approach is used
to analyze the collected data in order to develop overall daily trip generation models, as well as trip generation
models by trip purpose and time of day. Most of the trip generation models that were developed have acceptable
coefcient of determination (R
2
), while some of the models did not perform as expected. The Education Trips
model resulted in the highest R
2
, while the Off-Peak Trips model resulted in the lowest R
2
. Most of the inde-
pendent variables were also relevant to present the expected socioeconomic characteristics. Using land use
variables did not signicantly contribute towards improving the regression models, where only marginal im-
provements are noticed. It is suggested to investigate the transferability of the developed models to other Pal-
estinian cities, and to verify the applicability of these models to other cities with different sizes.
1. Introduction
Transportation planning processes and procedures are widely used to
estimate future trafc and travel demands. Estimated travel demands
are used to plan and assess future transportation facilities and services.
Travel demands are estimated based on various predictors that reect
socioeconomic and demographic characteristics. Planners tend to
quantify travel demand estimation based on these sets of predictors.
Travel demand estimates can be aggregated at different levels, such as
the individual or household levels, or can be formulated as an activity-
based, land use or zonal-based.
Developing travel demand models requires extensive efforts and re-
sources to collect, process and validate relevant data to reect both
travel behavior and socioeconomic predictors. Once reliable travel de-
mand models are developed, they can be used to estimate short–term
and long–term future trafc demands. However, the estimation of such
models in developing countries is faced with different challenges, such
as economic and political challenges, where projects prioritization and
logistics perspectives are key factors in decision making. Nonetheless,
travel demand models are still important to meet transportation stra-
tegic plans and studies for both developed and developing countries.
Since the establishment of the Palestinian National Authority (PNA)
in 1994, and as an emerging country with limited nancial resources
and political power, there is a persistent need to develop strategic
transport plans that are suitable to the current conditions and meeting
the anticipated mobility challenges (Almasri et al., 2010; Abu-Eisheh
and Ghanim, 2013a; Ghanim and Abu-Eisheh, 2013; Abu-Eisheh and
Ghanim, 2013b; Ghanim and Abu-Eisheh, 2015; Abu Zarifa and Sarraj,
2017; Abu-Eisheh and Ghanim, 2022). The recently prepared Road and
Transport Master Plan for Palestine had developed trip generation
model (The European Investment Bank and the Ministry of Trans-
portation and Systematica, 2016). However, the developed model was
built at the national level and does not meet the needs of local
communities.
The Palestinian cities have experienced considerable increase in
trafc demand, due to changes in transport policies and population
* Corresponding author.
E-mail address: mghanim@mot.gov.qa (M.S. Ghanim).
Contents lists available at ScienceDirect
Transportation Research Interdisciplinary Perspectives
journal homepage: www.sciencedirect.com/journal/transportation-
research-interdisciplinary-perspectives
https://doi.org/10.1016/j.trip.2024.101048
Received 19 October 2022; Received in revised form 20 September 2023; Accepted 15 February 2024
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
2
growth (Unviersal Group for Engineering and Consulting, 2012). This
increase is not in line with the anticipated transportation networks
improvement and infrastructure rehabilitation programs (Abu-Eisheh
et al., 2017). Accordingly, these factors (among others) are stressing the
need to develop localized travel demand estimation models. Such
models should be derived from the local needs to adapt to the economic,
social, and political challenges.
There is a signicant lack of specialized transportation planning
studies that are aiming to quantify, model, and estimate travel demand
forecasts in Palestinian cities because of political, economic, and social
constrains. Such studies are needed so that existing and future urban
transportation systems can accommodate the considerable increase in
trafc demand. Travel demand increase (either within or between the
Palestinian cities) is challenged with limited transport systems capacity
and noticeable high congestion levels. Thus, the need for developing
household-based trip generation models is persistent and urgent. Such
models would provide thorough understanding of travel demand needs
in the future. They will facilitate examining and evaluating the effect of
anticipated changes in socioeconomic characteristics and policies on
travel behavior and travel demand. Typically, estimating generated trips
is considered as the rst step in the widely adopted 4-step urban
transportation planning. Therefore, developing local trip generation
models for Palestinian cities is a key factor in improving the status of
transportation systems.
The current landscape of transportation planning in Palestinian cities
reveals a conspicuous deciency in specialized studies aimed at the
quantitative modeling and prediction of travel demand. This deciency
can largely be attributed to multifaceted political, economic, and social
constraints. It is imperative to address this gap as it has direct implica-
tions for the present and future viability of urban transportation systems,
particularly in light of the anticipated surge in trafc demand.
Furthermore, the rapid increase in urban trafc demand in Palestinian
cities, coupled with limited infrastructure improvements, poses a sig-
nicant challenge. Thus, developing localized models can be benecial
to local authorities, because of the constraints in adequately accom-
modating the increasing demand within the existing transportation
networks.
This escalating travel demand, whether occurring within the con-
nes of Palestinian cities or traversing between them, is compounded by
critical limitations in the capacity of existing transportation infrastruc-
ture and the concurrent prevalence of signicant congestion issues. As
such, there exists a pressing and persistent imperative to develop
household-based trip generation models. These models represent a
foundational framework for gaining a profound insight into the forth-
coming travel demand prerequisites.
The study is a comprehensive city-wide trip generation study that
was done within the Palestinian territories. This implies that there is a
lack of existing research on trip generation in Palestinian cities. The
study will address this gap by developing trip generation models for a
medium-size Palestinian city, the City of Jericho. It considers mixed land
use while developing the trip generation models, by investigating the
impact of mixed land use on trip generation. Moreover, the study focuses
on trip generation for trips with different purposes, such as educational,
shopping, social, and recreational purposes. This is a detailed approach
that is not widely adopted for developing cities in emerging countries,
where most of the trip generation models typically focused on overall
daily or peak-hour trips. Moreover, these models will facilitate a
comprehensive evaluation of the ramications stemming from envis-
aged changes in socioeconomic variables and policy dynamics on travel
behavior and demand patterns. It is crucial to underscore that the esti-
mation of generated trips constitutes the inaugural phase within the
widely adopted 4-step urban transportation planning methodology. In
light of these considerations, the imperative to formulate localized trip
generation models specically tailored to Palestinian cities assumes a
pivotal role in elevating the overall efcacy and sustainability of
transportation systems within the region.
In summary, this study addresses a substantial gap in the realm of
localized trip estimation models at the household level within Palestine.
Its primary objective is to close this gap by meticulously outlining the
developmental process of household-based trip generation models.
Moreover, this research endeavor involves a comprehensive descriptive
analysis of various demographic and socioeconomic predictors, shed-
ding light on their signicance in shaping travel patterns. Furthermore,
the study aims to devise practical and easily applicable models tailored
to the specic purposes and time of day for household-generated trips,
thereby enhancing the precision and usability of the resulting models.
2. Literature review
Several studies have documented the development of trip generation
models for different applications. Within the Palestinian territories,
Moussa conducted a study to develop a trip generation model for the
City of Gaza to determine the household travel characteristics pattern in
the study area (Moussa, 2013). In this study, multiple linear regression is
used to develop a model for attracted trips. They also compared the
conventional cross-classication against multiple cross–classication
for the City of Gaza. Household surveys were used to collect socioeco-
nomic data. A sample of 425 households from the different districts
within the city was considered. The results revealed that multiple
cross–classication models are more efcient in estimating trip rates
than the conventional cross-classication ones. Al-Sahili et al. developed
trip generation rates for major land uses in Palestinian cities using
regression analysis (Al-Sahili et al., 2018). These rates were the rst
attempts of estimating generated trips based on data collected from
specic land uses in those cities. More efforts were also made to develop
parking rates using regression analysis (Al-Sahili and Hamadneh, 2016).
Abu-Eisheh & Irshaid compared adaptive neuro-fuzzy inference systems
(ANFIS) and multiple linear regression (MLR) method in estimating
household-based generated trips for the City of Salt (Abu-Eisheh and
Irshaid, 2020). They found that ANFIS has shown promising results in
estimating generated trips when compared to MLR, with lower error and
variance. However, the sample size was based on 256 households.
As for other studies within the region, a cross-classication trip
generation rates for the City of Irbid in Jordan were developed (Al-
Masaeid and Fayyad, 2018). The study used regression analysis to esti-
mate home-based trip rates for work trips and other trips, along with
non-home-based trips. The study used three trip categories (household
size, income level, and vehicle ownership). Regression analysis
approach was used to develop trip generation rates for hospitals (Naser
and Faris, 2015) and shopping centers (Al Shehab and Khedaywi, 2018)
in Jordan.
Zonal-based trips for the City of Alexandria in Egypt were developed
using regression analysis (ELKAFOURY, n.d.). The study developed
linear models to estimate overall generated and attracted trips from each
of the 13 zones of Alexandria City. The study used the zonal population,
percentage of illiterates’ population, employment, age group between
25 and 35 years old, and educational land use percentage as explanatory
variables (or predictors).
Person trip generation rates for six different land uses in the City of
Dhaka, the capital of Bangladesh, were developed in one study (Ahmed,
2020). In-person interviews were used to estimate the modal share of
those person trips. They interviewed individuals who entered 20
different buildings and facilities that represented the selected land uses.
In addition to the trip rates, the study determined the passenger-car
equivalency factors for each of the eight different modes considered in
the study. Household-based trip generation rates for the City of Lipa,
Philippine were developed using regression analysis and trip rate
methods (Aloc and Amar, 2013). The study revealed that both methods
were acceptable to estimate household trips.
Another example of trip generation model from a nearby developing
countries is a study from Iraq. A statistical model was developed to
predict daily household trips based on socioeconomic data for
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
3
Al–Diwaniyah City (Soa et al., 2012). They used stepwise regression
method to develop the statistical mode. The city was divided into ve
sectors with 70 zones, and home-based surveys were distributed. The
results showed that the household generated trips can be estimated
based on household size, gender, number of employed individuals, and
number of students.
In another study in Iraq, statistical models were used to estimate
volumes of trips for a given design year for Al-Karkh area within the City
of Baghdad (Sarsam and Al-Hassani, 2011). The model considered both
motorized and non–motorized trips. The used trip rates per household
type to estimate generated trips volumes. Therefore, households were
classied according to their social class, income, and number of owned
vehicles. The results showed that total household trips are correlated
with household size, the number of male individuals, number of
employed individuals, number of students, and number of owned
vehicles.
Residential trip attraction model is developed for Dohuk City in Iraq
(Al-Taei and Taher, 2006). The city was divided into 28 different trafc
analysis zones located within the urban area, and the model considered
only 20 of these zones. The selected 20 zones hosted more than 300,000
residents. Travel data were collected using home-based interviews.
Travel data along with spatial data were used to develop the trip
attraction model.
Many studies have quantied trip generation in the context of
developed countries, through the estimation of proper representative
models. For instance, Guiliano et al. compared the travel behavior in two
developed countries, the United States of America, and the United
Kingdom. In their study, the results revealed that American participants
made on average 4.4 trips per day with an average travel distance of
approximately 31 miles (49.6 km), while the British participants made
3.0 trips per day for an average distance of 16 miles (25.6 km) (Giuliano
and Narayan, 2003; Giuliano, 2003; Giuliano and Dargay, 2006). Other
studies that took place in developed countries have used more advanced
techniques than linear regression methods, such as articial neural
networks, structural equations and cumulative logistic regression to
estimate different levels of generated trips (Amavi, 2014; Chang, 2014;
Cheng, 2019; Etu and Oyedepo, 2018; Ghanim and Ghassan, 2019;
Ghanim et al., 2013; Huntsinger et al., 2013; Kabakus¸ and Tortum,
2019).
Previous studies show that a range of socioeconomic, as well as land
use variables, inuence trip generation (De Gruyter et al., 2021). Some
of the variables are household size, age, gender, number of adults and
children, average household income, employment status, number of
employed individuals, educational status, vehicle ownership, and type
of dwelling unit. The effects of the built environment where sometimes
included in trip generation models, however, in a number of studies, it
was found that such effects are rather limited, as presented for example
by Ewing et al. (Ewing et al., 1996) and Boarnet & Sarmiento (Boarnet
and Sarmiento, 1998).
Literature reviews have also shown that regression analysis is widely
used as an acceptable tool to develop trip generation models for different
purposes, including freight trafc (Button, 1976; Veerappan, 2020; Sahu
and Pani, 2020; Balla et al., 2021). It was noticed that the use of
regression models is more common in countries and regions with
developing economies when compared to developed countries. This
approach is often enhanced by adopting the cross-classication method.
Regression analysis can be implemented at two levels. The rst level is
performed by correlating generated trips to the average zonal socio-
economic and demographic characteristics. This approach uses the
average zonal variables as input variables to estimate the number of
trips per average household in this zone, which is the output variable.
The second level uses more disaggregated data representing the house-
hold or individual characteristics as inputs, where the dependent vari-
able is the average number of trips per a given household or per person.
However, the selection of either level depends on the size of the study
area, homogeneity of socioeconomics, and demographic characteristics
of households.
3. Study area
The City of Jericho is chosen as a case study of urban areas in the
West Bank, Palestine. The city is located at the mid–eastern part of the
West Bank as shown in Fig. 1. The City is one of the oldest cities in the
world, where it was established more than 10,000 years ago. Currently,
it is the only access point for Palestinians living in the West Bank to
travel abroad through Jordan via the King Hussein’s Bridge located east
of the city. It is also an attraction point for tourism due to the valuable
historical places located within the city limits as well as its unique
weather, located at an altitude of about 260 m below mean sea level.
The city considered in this study can be classied as a medium-sized
urban area with a total geographic area of 57.43 km
2
, where part of
this area is located outside the ofcial municipal limits (Palestinian
Central Bureau of Statistics (PCBS), 2016). The city is one of the fastest
growing cities in Palestine, as it has attracted investors to develop eco-
nomic, touristic, recreational, and residential projects since the estab-
lishment of the PNA in 1994. It was the rst city in the West Bank to be
handed for the Palestinian administration after signing Oslo Accords in
1993.
According to the Palestinian Central Bureau of Statistics (PCBS), the
population of the City of Jericho and its vicinities was estimated as
35,885 inhabitants, where 23,822 inhabitants living in the city and its
suburbs, and 12,063 inhabitants who are living in adjacent refugee
camps (Palestinian Central Bureau of Statistics (PCBS), 2016). On the
other hand, the number of households in the city itself (i.e., the urban
part of the study area) was 3,510 who are living in 3,386 buildings with
4,549 residential units.
The study area was divided into 14 different Trafc Analysis Zones
(TAZs), based on several factors to identify those TAZs. Some of the
factors are the existing and expected land uses, distribution and densities
of population, and the existing natural and physical boundaries.
In terms of localities, the study area can be divided into three clus-
ters: the urban, the suburban/rural and the refugee camps. It is inter-
esting to indicate that having these three types of localities are typical
for the cities within the Palestinian territories (Palestinian Central Bu-
reau of Statistics (PCBS), 2016). The urban type is the largest, with a
population of 24,411, representing approximately 59.7 % of the total
population. The second largest type is the refugee camps, with a popu-
lation of 12,063 (33.6 %). The third type, with a suburban/rural nature,
has a population of 2,411 (6.7 %). Table 1 illustrates the diverse local-
ities included in the study area.
4. Sample framework and data preparation
Stratied and random household samples are drawn from the study
area. The stratication process covered all the geographic locations as
dened by the 14 TAZs. All of the randomly selected households were
visited, and personal interviews were conducted to collect demographic,
socioeconomic and travel behavior information. The face-to-face per-
sonal interviews method was chosen since this method is the most
effective in generating the highest response rate (Inc, 1996). The
response rate in this study was 100 % and results in the most precise and
accurate information. To reach this high response rate, the research
team consisted of two researchers: one male and one female from the
local community. The two team members identied their identity and
explained the signicance of the survey. Having two team members with
different genders was an effective strategy to gain the trust of the
households and encourage them to answer the survey questions. In
general, the Palestinians appreciate scientic contribution and promote
hospitality. The team members were welcomed to the interviewed
households. The households were cooperative and supportive, which
greatly contributed to the success of the surveys data collection mission.
The collected socioeconomic data include the household size, dwell
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
4
unit type, household members gender, education level, age distribution,
numbers of household employed persons, income, vehicle ownership,
motorcycle and bicycle ownership, and accommodation type. In this
research, the generated trips are divided into three categories. The rst
category is an overall aggregated trip generation model for the number
of generated trips per household. The second category includes different
trip generation models based on trip purposes. The last category in-
cludes trip generation models for different periods of the day. Table 2
summarizes and describes the independent variables, which were
identied to be of potential use in constructing trips estimation models.
On the other hand, trips information includes trip purpose, frequency,
timing, destination, and modal choice.
With respect to the sample size, it is important to use a sample size
that provides a statistically acceptable condence level. Since the pop-
ulation is known, the sample size based on nite population is initially
suggested (Sachs, 2012; Shaaban and Kim, 2016; Montgomery and
Runger, 2018). The minimum sample size for nite population (S) is
determined based on the formula shown in Eq. (1), as follows:
S=Ss
1+(Ss−1)
Pop
(1)
where:
S: Sample size for nite population
Ss: Sample size required for innite population
Pop :The population of the study area.
In order to estimate S, the sample size for the innite population, Ss,
is needed, which can be calculated as follows:
Fig. 1. The Location of the City of Jericho in the West Bank, Palestine.
Table 1
Localities in the Study Area and Estimated Population (Palestinian Central Bu-
reau of Statistics (PCBS), 2016).
Locality Name Locality Statistical
Code
Locality Type Population
Jericho City 351,920 Urban 21,411
Al New’ma 351,840 Suburban/
Rural
1,453
Ein Dyok Alfoqa 351,845 Suburban/
Rural
958
Ein Al Sultan Refugee
Camp
351,865 Refugee Camp 3,688
Aqbat Jaber Refugee Camp 351,975 Refugee Camp 8,375
Total urban population 21,411
Total suburban/rural
population
2,411
Total refugee camps
population
12,063
Total of Study Area
population
35,885
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
5
Ss=Z2
α
/2×P× (1−P)
C2(2)
where:
Z
α
/2: Normal statistic corresponding to
α
/2 signicance level.
P: Proportion of population making the selection, in decimal.
C: The desired condence interval or acceptable error margin, in
decimal.
In this study, it was assumed that the required condence level is 95
%, and ±0.02 margin of error is acceptable, and 50 % proportion, it was
estimated that the minimum number of households is 477, which will
meet the anticipated statistical condence in the results.
The sample size derived from the above (i.e., 477 households) was
compared against the U.S. Bureau of Public Roads (BPR) requirement,
which states that for population with less than 50,000 inhabitants, a
proportion of 10 % of the dwelling units can be examined (U.S. Bureau
of Public Roads, 1967). This can be translated into a sample size of 713
household. Since the BPR sample size was higher than the sample size
for a nite population, the targeted sample size in this study was 713
households.
5. Descriptive results
1.1. Household demographic and socioeconomic characteristics
The collected data were qualitatively processed and summarized to
provide a comprehensive perspective of the demographic, socioeco-
nomic and travel characteristics of the 713 surveyed households (508
independent units and 205 apartments). Table 3 summarizes different
household characteristics, such as household size distributions, number
of males and females in each household size group, average number of
employed individuals or students, vehicle ownership and the number of
individuals who have driving license.
5.1.1. Household size
The provided data are aggregated based on household size. Only 7
households have one individual, which represents less than 1 % of the
surveyed households. The largest surveyed households in terms of size
have 10 individuals. However, only three households of this size were
presented in the collected data. In general, the shape of household size
distribution follows a bell–shape, as shown in Fig. 2. The Shapiro-Wilk
normality test was performed, where H
0
implies that the household
size is normally distributed. The normality test result revealed that H
0
is
rejected, since the p-value is less than the signicance level of 0.05.
Table 4 summarizes the results of this test.
The surveyed households have a total population of 3,349 people
living in them, with an average of 4.7 individuals per household. It was
also found that 49.4 % of the individuals are male (1,655 males), and
51.6 % are females (1,694 females). In general, more than 50 % of
households are female, except for households of size 5 and 8.
5.1.2. Occupation
After discussing the key demographic characteristics of the surveyed
households, the socioeconomic characteristics are highlighted. In terms
of occupation, two main occupations are considered in this study,
employed individuals and students. The rst occupation is generating
income for the household, while the second one is purely for seeking
education. As can be seen in Fig. 3, most of the households have one or
two employed individuals, regardless of the household size. As for the
students, a proportional trend is seen, where the number of students in
the household increases as the household size increases. This observa-
tion is consistent with the age group distribution, where the age of
almost one-third of the individuals is less than 16 years, which is an age
group that is often highly correlated with students and education.
5.1.3. Vehicle ownership
With respect to vehicle ownership, Table 3 shows the average
number of owned vehicles per household based on its size. The general
trend is that as the household size increases, the vehicle ownership rate
increases, reaching the highest rate when the household size is 7, before
the trend starts to decrease again. The only exception was when the
household has a size of 10, which was presented by only three house-
holds, where each household in this category has exactly one vehicle. At
the household level, approximately half of the surveyed households do
not have any vehicle. As for those who do have at least one vehicle, 90 %
of them (46.1 % of the total households) have one vehicle, and most of
the remaining households have two vehicles, with only three households
with three vehicles. On a relevant matter, Table 3 shows the average
number of individuals per household based on its size who have valid
driver license. The observed trend is remarkably similar to the trend
seen for vehicle ownership. The relationship between owning a vehicle
Table 2
Independent and Dependent Variables and their Descriptions.
Variables Description
Demographic and Socioeconomic Inputs
HHSIZE Number of individuals living in household
MALE Number of male individuals in household
FEMALE Number of female individuals in household
EMP Number of employed individuals in household
EDU Number of students in household
AGE_A Number of individuals under the age of 16 years in household
AGE_B Number of persons between the age of 17 and 30 years in
household
AGE_C Number of persons between the age of 31 and 50 years in
household
AGE_D Number of persons between the age of 51 and 64 years in
household
AGE_E Number of persons above the age of 65 years in household
LICN Number of individuals with driver license in household
AUTO Number of owned vehicles in household
BIKE Number of owned bicycles in household
MOTO Number of owned motorcycles in household
INC Monthly household income (Thousands New Israeli Shekel)
HHTYPE Household dwell unit type (1 if Independent, 0 if Apartment)
Land Use Inputs
TAZ_SIZE Area of Trafc Analysis Zone, in squared kilometers
TAZ_Unit Number of households in the trafc analysis zone, in 100
′
s
TAZ_Pop Population of the trafc analysis zone, in 1,000
′
s
LU_RES Nominal variable indicating Residential TAZ land use (1 if exist,
0 if not)
LU_COM Nominal variable indicating Commercial TAZ land use (1 if exist,
0 if not)
LU_PUB Nominal variable indicating Public Facilities TAZ land use (1 if
exist, 0 if not)
LU_AGR Nominal variable indicating Agricultural TAZ land use (1 if exist,
0 if not)
LU_HTG Nominal variable indicating Cultural Heritage TAZ land use (1 if
exist, 0 if not)
LU_PKG Nominal variable indicating Parking TAZ land use (1 if exist, 0 if
not)
LU_IND Nominal variable indicating Industrial TAZ land use (1 if exist,
0 if not)
LU_PRK Nominal variable indicating Parks TAZ land use (1 if exist, 0 if
not)
Type of Trips
Daily Trips Average household’s daily trips rate
Work Trips Average household’s daily work trips rate
Education Trips Average household’s daily educational trips rate
Shopping Trips Average household’s daily shopping trips rate
Social Trips Average household’s daily social trips rate
Recreational
Trips
Average household’s daily recreational trip rate
Morning Trips Average household’s daily morning trips rate (before 9:00 AM)
Off-Peak Trips Average household’s daily off-peak trips rate (between 9:00 AM
– 4:00 PM)
Evening Trips Average household’s daily evening trips rate (after 4:00 PM)
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
6
and having a driver license for each household based on its size is
illustrated in Fig. 4, which shows a linear proportional trend.
As for the ownership of other modes of transportation, the results
show that 98 households own only one bicycle, 14 households own two
bicycles, only one household owns three bicycles, while the rest (600
households) do not own bicycles. On the other hand, motorcycle
ownership was much less, where only four households (out of 713) re-
ported owning only one motorcycle.
5.1.4. Age groups
In terms of age characteristics, Fig. 5 shows the age distribution of
the surveyed households. The gure shows that the age distribution is
skewed to right, with approximately one-third of the individuals are less
than 16 years old, and another one-third are between 17 and 30 years
old. These results are consistent with the national data published by the
Palestinian Central Bureau of Statistics (PCBS) (Palestinian Central Bu-
reau of Statistics (PCBS), 2023), which states that in 2023, approxi-
mately 35.2 % of the West Bank’s population is less than 14 years old.
The PCBS also states that 28.1 % are between 15 and 29 years old, and
only 6.3 % of the population is older than 60 years.
5.1.5. Household income
The sum of the monthly incomes of the surveyed households was
approximately 2,765,000 Shekel (approximately $700,000.00), with an
average monthly income of 3,878 Shekel (about $982) per household.
Fig. 6 summarizes the income distribution of the different households.
The gure shows that the income distribution is skewed to the right.
Approximately, 85 % of the surveyed households have an income of
5,000 Shekel ($1,266 or less). At the individual level, the average
monthly income is 826 Shekel ($209) per person.
1.2. Trafc analysis Zones and land use characteristics
The number of households in the city itself was estimated to be 3,510
living in 3386 buildings containing 4549 residential units, as indicated
by the PCBS. It stated that approximately 84.4 % of these buildings are
owned by their residents. The study area is extended over the city center,
the old residential area, areas with moderate density, newly constructed
suburbs, nearby rural areas, and refugee camps. Each of these areas is
characterized differently from others based on the land use, number of
households and housing units, constructed buildings, and population
density.
The study area was divided into TAZs. Different factors were
considered to identify those TAZs. Some of factors are the existing and
expected land uses, distribution and densities of population, and the
existing natural and physical boundaries (Cit of Jericho, Master Plan for
the City of Jericho, 2010). In total, the study area was divided into 14
different TAZs, as described in Table 5.
Table 3
Summary of Households’ Demographic and Socio-economic Characteristics.
Household
Size
Frequency Male
Frequency
Female
Frequency
Household Average No. of
Employed Persons
Household Average No.
of Students
Average No. of
Vehicles Owned
Average No. of Individuals
with Driver License
1 7 2 5 0.29 0.00 0.00 0.14
2 75 71 79 1.07 0.12 0.27 0.64
3 93 139 140 1.31 0.23 0.35 0.92
4 122 234 254 1.44 1.03 0.56 1.02
5 195 503 472 1.52 2.23 0.71 1.10
6 140 400 440 1.59 3.09 0.69 1.23
7 51 170 187 1.65 3.71 0.92 1.51
8 20 94 66 1.80 3.95 0.55 0.90
9 7 28 35 1.57 5.00 0.43 1.14
10 3 14 16 3.00 5.00 1.00 2.33
0
5
10
15
20
25
30
12345678910
Household Size Distriubtion, %
Household Size
Fig. 2. Household Size Distribution.
Table 4
Shapiro-Wilk Normality Test for Household Size Distribution.
Parameter Value
P-value 1.554e-13
W 0.9573
Sample size (n) 713
Average (x) 4.6971
Median 5
Sample Standard Deviation (S) 1.6449
Sum of Squares 1926.5638
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
7
1.3. Household-based trips characteristics
The 713 household surveys have resulted in a total of 4,890 trips for
an overall household size of 3,349 persons. This can be translated into an
average of 1.46 of daily trips per person. The estimated average trip rate
per person is compared against household trip rates from other inter-
national and regional studies, covering both well-developed and
developing economies. The average daily trips per person for different
Fig. 3. Number of Employed and Student Individual per Household.
y = 1.7199x + 0.1532
R² = 0.8326
0
0.5
1
1.5
2
2.5
0 0.2 0.4 0.6 0.8 1 1.2
Driver Licensed Rate
per Person per Household Size Group
Vehicle Ownership Rate per Person per Household Size Group
Fig. 4. Relationship between Vehicle Ownership and Number of Licensed Individuals.
32.4%
33.0%
27.0%
5.9%
1.7%
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
35.0%
Less than 16 17-30 31-50 51-64 65 or More
Percentage of Age Group, %
Age Group (Years)
Fig. 5. Overall Age Distributions for Surveyed Households.
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
8
regions and countries can be found in Table 6. The generated trips per
person derived from the study area is at the lower side when compared
with other regional and international rates. This nding can be associ-
ated with several factors, including household size, car ownership,
household economic and employment status, and social/cultural
perspectives.
The travel needs of the surveyed households can be classied by
either the purpose or the time of the trip. As for trip purpose, Fig. 7
shows ve different main purposes of trips were considered (Work,
Education, Shopping, Social and Recreation). The gure shows that
Education trips per household have the highest rate across the different
household sizes. Work and Shopping trips are following similar trend,
and they are close to each other, especially when the household size is
six or less. The number of Social trips does not have signicant variation,
regardless of the household size. Initiating Recreational trips has low
priority when compared to the other purposes. However, as the house-
hold size exceeds ve individuals per household, the number of recre-
ational trips increases and reaches a level that can be compared with
Work and Shopping trips. The number of Recreational trips drop as the
household size is exceeding eight individuals per households.
On the other hand, Fig. 8 shows three different time periods were
considered (AM peak, Off-peak, and PM peak). The gure shows that the
AM peak trips are the highest, especially if the household size is more
than four. The other observation is that the AM peak trips per household
are steadily increasing as the household size increases. While the PM
peak trips are increasing to a given limit, before they drop, where the
highest trips are associated with household size of eight. As for the off-
peak trips, the average number of trips is less than one during this
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
<1 1 - 2 2 - 3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 >10
Percentage of Households
Household Income (in Thousands of Shekel)
Fig. 6. Overall Income Distributions for Surveyed Households.
Table 5
Land Uses and basic Characteristics of Trafc Analysis Zones (Palestinian Central Bureau of Statistics (PCBS), 2016).
Trafc Analysis
Zone ID
Land Use Type Area*,
Km
2
Population Number
ofHousing Units
Residential Commercial Public
Facilities
Agricultural Cultural
Heritage
Parking Industrial Parks
1 ● ● ● 0.547 946 188
2 ● ● ● ● 2.479 2458 489
3 ● ● ● 2.409 958 190
4 ● ● ● ● ● ● 1.254 3,688 733
5 ● ● ● ● ● 3.726 1,453 289
6 ● ● ● ● ● ● 2.049 3256 647
7 ● ● ● ● ● ● 1.668 2816 560
8 ● ● ● ● 2.142 3047 606
9 ● ● ● 2.113 275 55
10 ● ● ● ● 4.185 1512 301
11 ● ● ● ● 1.262 2915 580
12 ● ● ● ● ● ● 3.864 814 162
13 ● ● ● ● ● ● 10.182 3179 632
14 ● ● ● ● ● 3.005 8,568 1703
Total 40.885 35,885 7134
*
Total area includes the inhibited areas and excludes agriculture dominated areas within the municipal limits.
Table 6
Daily Trip Generation Rate per Person derived from Regional and International
Studies.
Region, Country Daily Trip Rate
per Person
Source
Jericho, Palestine 1.46 (This Study)
United States of
America
3.37 (McGuckin & Fucci, 2018) (McGuckin and
Fucci, 2017)
Abu Dhabi, UAE 2.71 (UAE Ministry of Transport, 2015) (UAE
Department of Transport et al., 2015)
London, United
Kingdome
2.52 (WSP, 2014) (WSP, 2014)
Singapore 2.53 (Hou & Moogoor, 2019) (Hou and
Moogoor, 2019)
Ireland 2.65 (De Gruyter, 2019) (De Gruyter, 2019)
Kumasi, Ghana 2.39 (Taky) (Priyanto and Friandi, 2010)
Irbid, Jordan 2.44 (Al-Masaeid & Fayyad) (Al-Masaeid and
Fayyad, 2018)
Dohuk, Iraq 1.61 (Al-Taei & Taher) (Al-Taei and Taher,
2006)
Alexandria, Egypt 1.06 (ELKAFOURY, n.d.)
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
9
period, regardless of the household size. Table 7 illustrates the
descriptive statistics of the key independent variables.
6. Multivariable regression results and discussion
1.4. Modeling approach and model structure
The collected data from the 713 respondents are processed, and
stepwise multivariable regression analysis is performed using SPSS
software (Cronk, 2019). In each step, the regression model is evaluated
based on different statistical tests in order to assess the signicance of
each independent variable in the regression analysis. The statistical tests
used are t-test, collinearity statistics, and ANOVA test (Cronk, 2019). As
for the t-test, independent variables with insignicant t–value (based on
95 % condence level) are excluded from the model. The collinearity
statistics are used to determine if two or more variables are intercorre-
lated and then exclude the least signicant one. The ANOVA test is used
to provide an overall assessment of the signicance of the nal group of
predictors (i.e., independent variables) to provide reliable estimate to
the dependent variable when used together.
Eq. (3) presents the mathematical formula for the multivariable
linear regression models developed in this study. In total, and as pre-
sented in Table 1, there are 9 trip generation models, 8 demographic
variables, 8 socioeconomic variables, and 11 land use variables.
TRIPT=β0+∑
j
i=1
βi(Xi)(3)
where:
TRIPT: The estimate trips per household for trip type.T
β0: The regression analysis constant (i.e., y-intercept).
βi: Multivariable regression coefcient for predictor.i
Xi:Numerical value of independent variable i (i.e., predictor i).
1.5. Multivariable linear regression models for average daily trips per
household (DAILYTRIP)
As previously illustrated in the study objectives, the impact of land
use variables is assessed by developing trip generation models with and
without the use of those variables. The trip generation models are rst
developed using only demographic and socioeconomic characteristics of
households. These models are then developed by introducing the land
use variables. A comparison is then made between the developed models
(with and without) the land use variables using different performance
measures.
First, the average daily trips MLR models are developed and
compared using the statistical approach and procedures discussed pre-
viously. The nal estimated average daily trips per household model and
the regression analysis results of this model are highlighted in Table 8.
0
1
2
3
4
5
6
12345678910
Average Trip Rate per
Household
Household Size
Work
Education
Shopping
Social
Recreation
Fig. 7. Average Trips per Household for Different Purposes.
0
1
2
3
4
5
6
7
8
9
1 2 3 4 5 6 7 8 9 10
Average Trip Rate per
Household
Household Size
AM Peak
Off Peak
PM Peak
Fig. 8. Average Trips per Household for Different Time Periods.
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
10
In general, using the land use variables has resulted in slightly better
performance when compared to the model that does not use this input
group. However, the performance differences between the two models
are marginal. It is also noted that the number of independent variables
(i.e., predictors) for the model with land use variables is more than the
ones associated with the model without land use. In addition to the land
use variables, more socioeconomic and demographic variables appeared
in Model 2, such as household size, number of senior citizens who are
above 55 years, and vehicle ownership rate per household. Therefore,
the amount of additional household data that is needed to achieve a
marginal improvement in the model performance is not recommended.
Consequently, it is recommended not to use land use variables as part of
estimating average household daily trips. Nonetheless, the additional
benet of using land use variables for the other eight types of trips shall
be investigated.
The predictors summarized in Table 8 suggest that the number of
daily trips per household is best described using different independent
variables. However, both models have used employment and education
variables as predictors, where both of the two variables are statistically
signicant.
With respect to using socioeconomics and demographic variables to
predict daily number of trips (i.e., Model 1 without land use data), the
predictors are related to the employment, age, education, and monthly
income. The regression analysis reveals that the number of daily trips is
directly proportional to the number of employed persons and those
receiving education within the household, as well as with the household
income. Moreover, the number of daily trips is also directly proportional
to the number of young (under 16 years old) and old (between 51 and
64) individuals. This nding is consistent with the fact that young in-
dividuals have higher chances to generate educational trips, while old
individuals are more likely to generate work trips. What is worth
mentioning is that the middle-age group did not appear in the regression
analysis model. This nding is associated and consistent with the po-
litical, social, and economic constrains in Palestine. For instance, this
middle-aged group of individuals suffers from high rate of unemploy-
ment. Moreover, females within this age group have even lower
employment probability.
On the other side, when the land use variables are considered along
with socioeconomics and demographic variables (Model 2), other pre-
dictors are shown in the regression model, such as household size,
vehicle ownership, where each shows a proportional trend with daily
trips. Furthermore, when the number of senior household members in-
creases, it tends to demote the number of daily trips, which is shown
Table 7
Key Independent Variables and Their Descriptive Statistics.
Variable Average Median Standard
Deviation
Minimum Maximum
Demographic & Socioeconomics Explanatory Variables
HHSIZE 4.70 5.0 1.64 1.0 10.0
MALE 2.32 2.0 1.22 0.0 7.0
FEMALE 2.38 2.0 1.21 0.0 7.0
EMP 1.46 1.0 0.72 0.0 4.0
EDU 1.88 2.0 1.59 0.0 7.0
AGE_A 1.52 1.0 1.38 0.0 7.0
AGE_B 1.55 2.0 1.27 0.0 6.0
AGE_C 1.27 1.0 0.81 0.0 3.0
AGE_D 0.28 0.0 0.52 0.0 2.0
AGE_E 0.08 0.0 0.30 0.0 2.0
LICN 1.06 1.0 0.89 0.0 5.0
AUTO 0.59 1.0 0.59 0.0 3.0
BIKE 0.18 0.0 0.44 0.0 3.0
MOTO 0.01 0.0 0.07 0.0 1.0
INC 3.88 3.5 2.16 0.8 20.0
HHTYPE Independent Household: 71.2 % – Apartment Household: 28.8 %
Land Use Explanatory Variables
TAZ_Area 2.92 2.28 2.25 0.55 10.18
TAZ_Pop 2.56 2.64 1.97 0.28 8.57
TAZ_Unit 5.10 5.25 3.92 0.55 17.03
LU_RES 100 % of Trafc Analysis Zones has Residential Land Use
LU_COM 85.7 % of Trafc Analysis Zones has Commercial Land Use
LU_PUB 85.7 % of Trafc Analysis Zones has Public Facilities Land Use
LU_AGR 92.9 % of Trafc Analysis Zones has Agricultural Land Use
LU_HTG 28.6 % of Trafc Analysis Zones has Cultural Heritage Land Use
LU_PKG 21.4 % of Trafc Analysis Zones has Parking Land Use
LU_IND 14.3 % of Trafc Analysis Zones has Industrial Land Use
LU_PRK 14.3 % of Trafc Analysis Zones has Parks Land Use
Number of Trips per Trip Type (Dependent Variables)
Daily Trips 6.89 7.0 3.10 0.0 16.0
Work Trips 1.60 1.0 0.82 0.0 5.0
Education
Trips
1.84 2.0 1.57 0.0 6.0
Shopping Trips 1.52 1.0 1.00 0.0 7.0
Social Trips 1.03 1.0 1.14 0.0 7.0
Recreational
Trips
0.92 0.0 1.45 0.0 6.0
Morning Trips 2.95 3.0 1.82 0.0 10.0
Off-Peak Trips 3.15 3.0 1.75 0.0 9.0
Evening Trips 0.77 1.0 0.80 0.0 5.0
Table 8
Results of Regression Analysis for General Trip Generation Model (Average Number of Household Daily Trips, DAILYTRIP).
Variable Without Land Use Variables (Model 1) With Land Use Variables (Model 2)
Coefcient Standard Error t-value p-value Coefcient Standard Error t-value p-value
Constant 1.82 0.184 9.915 0.000 0.977 0.345 2.834 0.005
EDU 1.35 0.058 23.039 0.000 1.193 0.074 16.016 0.000
EMP 1.29 0.098 13.208 0.000 1.062 0.102 10.389 0.000
AGE_A 0.20 0.068 2.907 0.004 — — — —
AGE_D 0.28 0.136 2.076 0.038 — — — —
INC 0.07 0.034 2.074 0.038 0.079 0.032 2.437 0.015
HHSIZE — — — — 0.224 0.071 3.140 0.002
AGE_E — — — — −0.617 0.217 −2.841 0.005
AUTO — — — —
TAZ_Pop — — — — 0.377 0.088 4.304 0.000
LU_HTG — — — — 0.508 0.172 2.949 0.003
LU_PKG — — — — −3.415 0.558 −6.122 0.000
LU_IND — — — — 3.035 0.542 5.594 0.000
R 0.831 R 0.850
F-value 315.590 F-value 203.138
Standard Error 1.721 Standard Error 1.642
R2 0.691 R2 0.722
Adjusted R2 0.688 Adjusted R2 0.719
Sample Size 713 Sample Size 713
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
11
with the negative sign in the regression model. Finally, land use pre-
dictors associated with income-generating activities (e.g., industrial
activities) are proven to promote daily trips.
1.6. Multivariable linear regression models for different trip purposes
Multivariable regression analysis (with and without land use vari-
ables) was performed for the other types of trips, which represent
different trip purposes and different times of day. Table 9 summarizes
the performance of the regression analysis models. In general, there is a
great variation of model performances regardless of the used inputs. For
instance, educational trips and morning trips models have the highest
accuracy and the lowest standard error. On the other hand, other models
have shown poor performance using MLR, such as shopping, social, off-
peak, and evening trips. The performances of these trips’ models were
not generally improved when land use variables were introduced to the
MLR models, where only marginal improvements are observed, if any.
Models that perform well without land use variables have slight im-
provements when land use variables are used. As for models with poor
performance without land use variables, their performance is still poor
when land use data are considered, despite the slight improvement.
The low performance of certain trips’ models indicates that there is a
need to collect other types of explanatory household survey questions. It
is also noticed that the models with low performance are reecting the
economic status of the surveyed households. With respect to the purpose
of trips for instance, shopping, social and recreational trips tend to
reect correlation with household nancial conditions. Therefore,
questions such as average monthly expenses, cost of transportation, and
dept conditions can potentially provide predictors to improve the per-
formance of those models. The nancial conditions could also play a
signicant role in improving the performance of the evening and off-
peak trips, since those trips might be considered secondary trips that
are not associated with educational or employment. This would also
explain why the model for morning trips has performed well.
Table 10 and Table 11 are summarizing the regression analysis
models’ coefcients and constants, along with the signicance of each
coefcient with and without the use of land use variables. For simplicity,
the results are presented in a tabular format. It is obvious that models
with land use variables show more predictors than the models that did
not consider land use when they were developed.
With respect to the nature of the predictors appearing in each model,
it can be stated that, in general, the independent variables used to
explain the variance in each of the dependent variables is logical. For
instance, work–related trips depend on employment. Educational trips
depend on the number of students. The educational trips model implies
that a household with working parents and young children is most likely
to generate educational trips.
It is noticed that shopping trips are not derived by the household
income. Instead, these trips are derived by the number of students and
employed individuals per household. This observation indicates that the
main reason behind shopping is to serve the basic needs for these two
groups, rather than shopping for nonessential commodities. This
observation can also be taken as a reection of the economic hardship
for most of the surveyed households.
For the social trips, it can be noticed that the number of young in-
dividuals per household is promoting this trip type. Moreover, as the
number of male individual increases, this type of trips tends to be
reduced. This observation is consistent with the Palestinian commu-
nities in general, where social trips tend to be a household activity to
entertain young individuals (mostly children). For recreational trips, the
income level, number of students, and parent’s age are associated with
generating this trip type. As for the impact of land use on social trips, the
model shows that the number of households in a trafc analysis is
inversely proportional to the number of social trips.
On the other hand, new independent variables associated with
generated trips based on time of day are different from those used for
trip purposes. For instance, the variable LICN expressing the number of
licensed drivers in the household appeared in both AM and PM peak
trips. However, the variable LICN is inversely proportional with the AM
peak trips. This nding is related to the cost of transportation in Pales-
tinian cities. Most of the adults who have secured jobs tend to obtain a
driving license, yet they tend to use public transportation. The opera-
tional cost of owning and running private vehicles is expensive and
unaffordable for many households. Not only that the gas price in the
Palestinian territories is high, but it is also inconsistent with the average
monthly household income (Karaeen, et al., 2013). Furthermore, the
different land uses are better presented on this type of trip.
1.7. The signicance of linear regression model constant
The models previously presented are behaving differently with
respect to the signicance of the linear regression model constant. While
most constants are statistically signicant, some of them are insigni-
cant. The ideal case for selecting the linear regression model is when the
constant and the coefcients are all statistically signicant. Therefore,
the signicance of the constants and the coefcients are important in
selecting the regression model. There are three possibilities for the
regression model formula. The rst case is when the constant and co-
efcients are statistically signicant. The second case is when the con-
stant is insignicant (i.e., there is no evidence that the constant is
statistically different than a value of zero). The third case is when the
constant is signicant, and the coefcient is insignicant. The stepwise
multivariable linear regression analysis models developed in this study
belong to the rst or second cases only. Several studies have investigate
the proper way of deciding whether or not to use the constant part of
linear model (Button, 1976; Sahu and Pani, 2020; Balla et al., 2021;
Holguin-Veras et al., 2002; Pani, 2018). The preferred model in this
study is the one with a constant (i.e., y–intercept). However, if the
constant turns out to be statistically insignicant (i.e., the constant is
associated with an insignicant t-value), then the model is reconsidered
to have no constant. Table 12 is presenting the multivariable linear
regression models without a constant only the models without a sig-
nicant constant (i.e., y-intercept). Table 13 provides a summary of the
performance of the regression models, which shows that removing the
Table 9
Summary of Regression Analysis Results for All Trip Generation Models.
Type of Trips Model Without Land Use Variables With Land Use Variables
R R2 Adj.R2 Std. Error R R2 Adj.R2 Std. Error
Daily Trips 0.831 0.691 0.688 1.721 0.850 0.722 0.719 1.642
Work Trips 0.857 0.734 0.734 0.421 0.858 0.736 0.736 0.419
Education Trips 0.984 0.969 0.969 0.277 0.984 0.969 0.969 0.277
Shopping Trips 0.327 0.107 0.105 0.940 0.096 0.009 0.004 2.409
Social Trips 0.182 0.033 0.032 1.123 0.277 0.077 0.076 1.097
Recreational Trips 0.325 0.105 0.103 1.370 0.374 0.140 0.138 1.344
Morning Trips 0.949 0.901 0.900 0.552 0.952 0.905 0.905 0.539
Off-Peak Trips 0.147 0.022 0.020 0.795 0.284 0.081 0.079 0.771
Evening Trips 0.473 0.223 0.218 1.602 0.513 0.263 0.260 1.561
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
12
constant from the model with insignicant constant has improved its
overall performance. It is recommended to use those models over ones
previously presented in Table 10 and Table 11.
1.8. Discussion
Multiple household-based trip generation models were previously
presented, covering different trip purposes. The models were developed
using data from a household travel survey conducted in the City of
Jericho, Palestine. The data included information on household char-
acteristics, socioeconomic factors, and travel behavior. This section
highlights the contribution of some independent variables that are
dominating the developed models.
Education: The level of education of the household head was posi-
tively associated with the number of trips generated for work and social
trips. This means that households with more educated heads tend to
generate more trips for these purposes. This is likely because higher-
educated people tend to have more job opportunities and social en-
gagements that require them to travel.
Household Size: The number of people in a household was posi-
tively associated with the number of trips generated, for all trip pur-
poses. This means that households with more people tend to generate
more trips. This is likely because larger households have more people
who need to travel for different purposes, such as work, school, and
shopping.
Employment: Household income was positively associated with the
number of trips generated for work and social trips. This means that
households with higher incomes tend to generate more trips for these
purposes. This is likely because higher-income households have more
disposable income, which they can use to travel.
Vehicle Ownership: The number of cars owned by a household was
positively associated with the number of trips generated for all trip
purposes. This means that households with more cars tend to generate
more trips. This is likely because cars provide a convenient and exible
way to travel, which makes it easier for people to make trips.
Household Age Distribution: The results show that households
with young individuals are more likely to be associated with Education,
Social, and Morning trips. This conclusion is intuitive, since this age
group is typically linked with school trips, especially in the morning,
where the vast majority of school starts at 7:00 am in the morning. The
interesting part however is the correlation between this age group and
social trips. This implies that households with young individuals are
more likely to perform social trips, for leisure purposes and for staying
connected with relatives and other extended family members. Further-
more, the results show the negative inuence of the old-age group and
different types of trips.
The results of the study can be generalized to other cities with similar
characteristics to the City of Jericho. However, it is important to
consider the specic characteristics of the city before applying Jericho’s
models. The results of the study can also be used to support trans-
portation planning and policy in other cities by establishing a bench-
mark for comparison or developing city-specic models.
There are some limitations associated with the developed models.
For example, the models were developed based on data collected from
one medium-sized yet fast growing city. The study only included data
from household travel surveys, so the results may not be generalizable to
all travelers, despite the fact that the sample size meets the international
best-practice guidelines. Another limitation is the considered factors,
where the impact of other factors can be investigated, such as detailed
land use, public transportation coverage, and transportation infra-
structure, on trip generation. Despite these limitations, the study pro-
vides valuable insights into the factors that inuence trip generation.
The results of the study can be used to improve transportation planning
and policy in the City of Jericho and other cities in Palestine.
Table 10
Summary of MLR Coefcients for Trip Generation Models (without Land Use Predictors).
Predictor Trips Type
Work
a
Education Shopping Social Recreational Morning Off-Peak Evening
Constant 0.148
b
3.701
c
| 0.000
d
−0.038
−1.358 | 0.175
0.896
0.000 | 0.000
0.963
10.215 | 0.000
0.214
1.726 | 0.085
0.033
0.499 | 0.618
0.696
14.378 | 0.000
1.209
7.368 | 0.000
EDU 0.022
2.204 | 0.028
0.947
88.406 | 0.000
0.185
0.000 | 0.000
—
e
0.217
5.5 | 0.000
0.902
41.729 | 0.000
— 0.357
6.562 | 0.000
EMP 0.968
44.23 | 0.000
— 0.193
0.000 | 0.000
— — 0.903
29.176 | 0.000
— 0.38
4.374 | 0.000
AGE_A — 0.025
2.371 | 0.018
— 0.155
4.871 | 0.000
— 0.127
5.925 | 0.000
— 0.169
2.651 | 0.008
AGE_B — — — — — — 0.064
2.717 | 0.007
—
AGE_C — 0.038
2.502 | 0.013
— — −0.2
−2.668 | 0.008
0.065
2.127 | 0.034
— —
AGE_D — — — — — — — 0.391
3.107 | 0.002
AGE_E — — — — — — −0.267
−2.643 | 0.008
—
AUTO — — — — — — — 0.262
2.365 | 0.018
FEMALE — — — — — −0.049
−2.168 | 0.031
— —
INC — — — — 0.142
5.717 | 0.000
— — —
LICN — −0.034
−2.792 | 0.005
— — — −0.054
−2.087 | 0.037
— —
MALE — 0.021
2.145 | 0.032
— −0.075
−2.079 | 0.038
— — — —
Note:
a- Each trip rate predictors are read vertically.
b- The upper number in the cell is the coefcient value.
c- The lower-left number in the cell is the t-value associated with this coefcient.
d- The lower-right number in the cell is the p-value associated with this coefcient.
e- Missing values indicate that variable did not appear in the corresponding model.
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
13
7. Conclusions and recommendations
This study highlights the efforts made to develop trip generation
models for a medium size developing city in the emerging country of
Palestine. The efforts include identifying and quantifying land use var-
iables, along with the key socioeconomic and demographic information
per households, along with the travel needs for these households. Travel
needs were expressed based on identifying different trip purposes and
times. A sample of 713 households from the City of Jericho was selected
in this study. Different trip generation models per household were
developed using multivariable linear regression analysis to correlate
numbers of generated trips to several socioeconomic and demographic
variables. There are three different groups for these models. One of the
models is an overall trip generation model, while the other two groups
cover the different trip purposes and time spans of the days. The results
of regression models for the different types of estimates trips have
different performance. For instance, the R2 values for these models
range between 0.022 and 0.734. The results also show that marginal role
for land use variables in estimating household trips.
Overall, the developed models can be used to enhance transportation
planning efforts for the Palestinian cities. These models are easy to
implement by transport planners and engineers. They provide realistic
estimates of trips at the household level. Their value can be further
appreciated due to the lack of trip estimation models that are derived
from the local needs to the Palestinian cities. Moreover, the study paves
the road to develop similar models for other cities or update them in the
future as the underlying prevailing conditions change.
The output of such studies on trip generation can be further used in
the next steps towards arriving at the travel demand reected by fore-
casted trafc volumes on the transportation system for the largely
established, yet rapidly growing cities. The local governments can use
these methods to supplement planned growth in the future. These can be
utilized in the preparation and development of future land use/trans-
portation studies and urban master plans.
The signicance of this study is its transferability potential. In gen-
eral, the Palestinian cities share similar demographic and socioeconomic
Table 11
Summary of MLR Coefcients for Trip Generation Models (with Land Use Predictors).
Predictor Trips Type
Work
a
Education Shopping Social Recreational Morning Off-Peak Evening
Constant 0.0100.136|
0.892
−0.038–1.358|
0.175
0.985
4.173|0.000
1.22512.867|
0.000
0.0480.349|
0.727
−0.061–0.433|
0.665
0.71510.963|0 0.4561.166|
0.244
HHSIZE — — 0.1154.045|
0.000
— — — — 0.2123.101|
0.002
EDU 0.0212.149|
0.032
0.94788.406|0 — — 0.2025.126|0 0.897
43.318|0.000
— 0.2393.392|
0.001
EMP 0.96944.393|
0.000
— — — — 0.882
30.493|0.000
— 0.2512.651|
0.008
AGE_A — 0.0252.371|
0.018
— 0.113.623|0.000 — 0.118
5.636|0.000
— —
AGE_B — — — — — — 0.094
4.022|0.000
—
AGE_C — 0.0382.502|
0.013
— — −0.196–2.642|
0.008
0.062.023|0.043 — —
AGE_E — — — — −0.36–2.032|
0.043
— −0.254–2.589|
0.01
—
AUTO — — — — — — — 0.2322.124|
0.034
FEMALE — — — — — −0.049–2.24|
0.025
— —
INC — — — — 0.1315.296|
0.000
— — —
LICN — −0.034–2.792|
0.005
— — — — — —
BIKE — — −0.18–2.314|
0.021
— — — — —
MALE — 0.0212.145|
0.032
0.0962.499|
0.013
— — — — —
TAZ_Area — — — −0.071–4.063|0 0.0864.009|0 −0.043–4.887|0 — —
TAZ_Pop — — −5.959–5.232|
0.000
— — −0.027–3.402|
0.001
−0.034–3.07|
0.002
—
TAZ_UNIT — — 3.0795.143|
0.000
— — — — —
LU_COM — — — — — — — 0.8282.471|
0.014
LU_PUB 0.1492.375|
0.018
— — — — — — —
LU_PRK — — — — — — 0.427
5.416|0.000
—
LU_HTG — — 0.2622.77|0.006 — — — — —
LU_PKG — — — −0.416–4.745|
0.000
— — — −0.683
−5.212|0.000
LU_AGR — — −0.538–2.276|
0.023
1.22512.867|
0.000
0.0480.349|
0.727
0.3482.651|
0.008
— —
Note:
a- Each trip rate predictors are read vertically.
b- The upper number is the coefcient value.
c- The lower-left number in brackets is the t-value associated with this coefcient.
d- The lower-right number in brackets is the p-value associated with this coefcient.
e- Missing values indicate that variable did not appear in the corresponding model.
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
14
characteristics, and it is expected that these models are applicable and
can be transferred to other cities. However, further studies are needed to
assure the application of transferability to other cities before conrming
the transferability potential. Although this study was conducted to
generate trips for a medium-size city, there are other cities that are
smaller and larger in Palestine. This study can be extended to those cities
after proper testing and calibration, which would also be benecial to
verify model transferability as discussed earlier. As an emerging country
with limited nancial resources and political inuence, Palestine may
encounter limitations in the availability of local resources to support the
development of sophisticated transportation planning models, poten-
tially impacting the study’s depth and scope.
While this study offers valuable trip generation models for a devel-
oping urban environment, it is essential to acknowledge certain limita-
tions that could serve as focal points for future research endeavors.
Notably, one of the key limitations inherent in this study pertains to the
intricacies associated with data collection, processing, and validation,
which are vital components in constructing accurate travel demand
models. This undertaking necessitates considerable resources and ef-
forts, and within the Palestinian context, it may encounter heightened
challenges due to prevailing economic and political constraints. Addi-
tionally, given Palestine’s status as an emerging nation with constrained
nancial resources and limited political inuence, there exists the po-
tential for limitations in the accessibility of local resources to facilitate
the development of advanced transportation planning models. These
constraints could potentially impinge upon the study’s breadth and
depth, warranting further exploration in future research initiatives.
Furthermore, travel behavior is a multifaceted phenomenon inuenced
by an array of factors, including social, economic, and political de-
terminants. While this study endeavors to capture and incorporate these
factors into its models, the complexity of human behavior suggests that
additional dimensions may be worth investigating in order to attain a
more comprehensive understanding of travel patterns. Moreover, it is
crucial to recognize the dynamic nature of the political, economic, and
Table 12
Summary of MLR Coefcients Trip Generation Models with Insignicant Constant (y-intercept).
Predictor Trips Type without Land Use Predictors Trips Type (with Land Use Predictors)
Education* Recreation* Morning* Work* Education* Recreation* AM* PM*
HHSIZE — — — — — — — 0.204
3.246|0.001
EDU 0.95494.874 | 0.000 0.1755.436 |
0.000
0.90042.609 |
0.000
0.0222.248 |
0.025
0.95494.874 | 0.000 0.203
5.155|0.000
0.88145.607 |
0.000
0.251
3.611|0
EMP — — — 0.970
49.925|0.000
— — 0.90340.537 |
0.000
0.238
2.55|0.011
AGE_A 0.0242.347 | 0.019 — 0.91234.867 |
0.000
— 0.0242.347 | 0.019 — 0.1216.068 | 0.000 —
AGE_C 0.0352.672 | 0.008 — 0.1296.079 |
0.000
— 0.0352.672 | 0.008 −0.185
−2.740|
0.000
0.0722.521 | 0.012 —
AGE_D — 0.2172.172 |
0.03
— — — — — —
AGE_E — — — — — −0.347
−2.005|
0.045
— —
AUTO — — 0.0692.375 |
0.018
— — — 0.294
2.728|0.007
INC — 0.1357.038 |
0.000
— — — 0.136
6.340|0.000
— —
LICN −0.033–3.103 |
0.002
— — — −0.033–3.103 |
0.002
— —
TAZ_Area — — — — — 0.089
4.543|0.000
−0.035–4.43 |
0.000
0.055
2.149|0.032
TAZ_Unit — — — — — — — 0.176
6.006|0.000
LU_IND — — — — — — — 2.35
6.211|0.000
LU_PUB — — — 0.1564.408|
0.000
— — —
LU_HTG — — — — — — — 0.417
2.67|0.008
LU_PKG — — — — — — — −2.693
−6.744|
0.000
*
Indicates that this regression model does not have a constant.
Table 13
Summary of Regression Analysis Results for All Trip Generation Models.
Type of Trips Model Trips without Land Use Variables Trips with Land Use Variables
R R2 Adj.R2 Std. Error R R2 Adj.R2 Std. Error
Work* — — — — 0.972 0.946 0.945 0.420
Education* 0.984 0.969 0.969 0.277 0.993 0.987 0.987 0.281
Recreation* 0.325 0.105 0.103 1.370 0.616 0.379 0.375 1.357
Morning* 0.949 0.901 0.900 0.552 0.988 0.977 0.976 0.553
Evening* — — — — 0.893 0.797 0.795 1.571
*
Regression model has no constant (i.e., y-intercept is 0.000).
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
15
social landscape in Palestine. These factors can evolve rapidly, poten-
tially exerting a transformative inuence on the applicability and rele-
vance of the developed models over time. The inherent difculty in
forecasting these external variables introduces an element of uncertainty
into the long-term implications of the study, prompting the need for
ongoing monitoring and adaptation of the models in response to
changing conditions.
The study focused on a medium-size city in Palestine, but it can be
expanded to include other cities of different sizes or to include other
factors that inuence travel behavior, such as the availability of public
transportation. It can also be expanded to verify that these models are
applicable to other cities of different sizes.
Although multivariable linear regression models are developed, the
use of other advanced modeling approaches can be developed, such as
machine learning models. These models can potentially improve trip
generation prediction accuracy as they are capable of efciently
modeling other nominal variables, such as zonal and land use charac-
teristics to improve the response of the dependent variables to their
predictors. Such an adoption of machine learning approaches will
encounter the need to collect large datasets and to cover wide factors
spectra. Moreover, further studies can investigate the development of
hybrid or nested models, where more than one step from the 4-step
transportation planning steps can be modeled at the same time.
The study also acknowledges that the political, economic, and social
landscape in Palestine is dynamic, and the models should be monitored
and adapted over time to reect changes in these conditions.
CRediT authorship contribution statement
Sameer Abu–Eisheh: Conceptualization, Methodology, Supervi-
sion, Validation, Writing – review & editing. Mohammad S. Ghanim:
Methodology, Data curation, Formal analysis, Writing – review & edit-
ing. Alaa Dodeen: Investigation.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgement
Open Access funding provided by the Qatar National Library.
References
Abu Zarifa, H.K., Sarraj, Y.R., 2017. Trafc Impact of Planned Gaza Seaport on Major
Roads in Gaza Strip, Palestine. Journal of Engineering Research and Technology 4
(3).
Abu-Eisheh, S.A. and M.S. Ghanim. Development and Evaluation of Trafc Systems
Management Course for Sustainable Built Environment in Developing Countries. in
Transportation Research Board 92nd Annual Meeting. 2013.
Abu-Eisheh, S., M. Ghanim, and A. Dodeen. Modeling travel behavior based on
household characteristics. in MATEC Web of Conferences. 2017. EDP Sciences.
Abu-Eisheh, S.A., Ghanim, M.S., 2013b. Managing Transportation for Sustainable Built
Environment By Developing A Trafc Systems Management Course. Procedia-Social
and Behavioral Sciences 102, 499–507.
Abu-Eisheh, S.A., Ghanim, M.S., 2022. Improving senior-level students’ performance in
Trafc Systems Management using multimedia contents. Ain Shams Engineering
Journal 13 (1), 101511.
Abu-Eisheh, S., Irshaid, M., 2020. Modelling Trip Generation using Adaptive Neuro-
Fuzzy Inference System in Comparison with Traditional Multiple Linear Regression
Approach. International Journal of Simulation-Systems. Science & Technology 21
(2).
Ahmed, T., et al., 2020. Trip Generation Rates of Land Uses in a Developing Country City.
Transportation Research Record 2674 (9), 412–425.
Al Shehab, O.M., Khedaywi, T.S., 2018. Trip and Parking Generation for Shopping
Centers in Jordan. Institute of Transportation Engineers. ITE Journal 88 (2), 45–49.
Al-Masaeid, H.R., Fayyad, S.S., 2018. Estimation of trip generation rates for residential
areas in jordan. Jordan. Journal of Civil Engineering 12 (1).
Almasri, E.H., Sarraj, Y.R., Eljamassi, A.D., 2010. Transportation Planning Challenges in
Developing Cities–A Practice from Rafah. Palestine.
Aloc, D.S. and J. Amar. Trip generation modelling of lipa city. in Seminar and research
methods in civil engineering research program, University of Philippines Diliman.
doi. 2013.
Al-Sahili, K., Hamadneh, J., 2016. Establishing parking generation rates/models of
selected land uses for Palestinian cities. Transportation Research Part a: Policy and
Practice 91, 213–222.
Al-Sahili, K., Eisheh, S.A., Kobari, F., 2018. Estimation of new development trip impacts
through trip generation rates for major land uses in Palestine. Jordan. Journal of
Civil Engineering 12 (4).
Al-Taei, A.K., Taher, M., 2006. Trip Attraction Development Statistical Model in Dohuk
City Residential Area. Al-Radain Engineering 14 (2), 12–23.
Amavi, A.A., et al., 2014. Advanced trip generation/attraction models. Procedia-Social
and Behavioral Sciences 160, 430–439.
Balla, B.S., Sahu, P.K., Pani, A., 2021. Are Freight Production Models Transferable
between Urban and Suburban Areas? Guiding Model Transfer in Geographically
Sprawling Indian Cities. Journal of the Institution of Engineers (india): Series A 102
(3), 643–656.
Boarnet, M.G., Sarmiento, S., 1998. Can land-use policy really affect travel behaviour? A
study of the link between non-work travel and land-use characteristics. Urban
Studies 35 (7), 1155–1169.
Button, K., 1976. Category Analysis & Household Multiple Regression Models of Trip
Generation: A Possible Reconciliation. Rivista Internazionale Di Economia Dei
Trasporti/international Journal of Transport Economics 19–27.
Chang, J.S., et al., 2014. Comparative analysis of trip generation models: results using
home-based work trips in the Seoul metropolitan area. Transportation Letters 6 (2),
78–88.
Cheng, L., et al., 2019. Structural equation models to analyze activity participation, trip
generation, and mode choice of low-income commuters. Transportation Letters 11
(6), 341–349.
Cit of Jericho, Master Plan for the City of Jericho. 2010: Jericho.
Cronk, B.C., How to use SPSS®: A step-by-step guide to analysis and interpretation. 2019:
Routledge.
De Gruyter, C., 2019. Multimodal Trip Generation from Land Use Developments:
International Synthesis and Future Directions. Transportation Research Record 2673
(3), 136–152.
De Gruyter, C., Zahraee, S.M., Shiwakoti, N., 2021. Site characteristics associated with
multi-modal trip generation rates at residential developments. Transport Policy 103,
127–145.
ELKAFOURY A., et al., Develop Trip Generation Model for Alexandria City-Egypt.
Journal of Fluids, Heat and Mass Transfer, Mechanical and Civil Enginnering.
Etu, J., Oyedepo, O., 2018. Forecasting Trip Generation for High Density Residential
Zones of Akure, Nigeria: Comparability of Articial Neural Network and Regression
Models. Journal of Civil Engineering, Science and Technology 9 (2), 2.
Ewing, R., DeAnna, M., Li, S.-C., 1996. Land use impacts on trip generation rates.
Transportation Research Record 1518 (1), 1–6.
Ghanim, M.S., Abu-Eisheh, S.A., 2015. The impact of random mid-block pedestrian
crossing on urban arterial operational characteristics in developing countries.
International Journal of Engineering Management and Economics 5 (3–4), 169–180.
Ghanim, M.S. and S.A. Abu-Eisheh. The impact of mid-block crossing on urban arterial
operational characteristics using multimodal microscopic simulation approach. in
2013 5th International Conference on Modeling, Simulation and Applied
Optimization (ICMSAO). 2013. IEEE.
Ghanim, Mohammad S., Ghassan, Abu-Lebdeh, Kamran, Ahmed, 2013. Modeling
historical trafc data using articial neural networks. In 2013 5th International
Conference on Modeling, Simulation and Applied Optimization (ICMSAO). IEEE,
pp. 1–4.
Ghanim, Mohammad S., Ghassan, Abu-Lebdeh, 2019. Projected state-wide trafc
forecast parameters using articial neural networks. IET Intelligent Transport
Systems 13 (4), 661–669.
Giuliano, G., 2003. Travel, location and race/ethnicity. Transportation Research Part a:
Policy and Practice 37 (4), 351–372.
Giuliano, G., Dargay, J., 2006. Car ownership, travel and land use: a comparison of the
US and Great Britain. Transportation Research Part a: Policy and Practice 40 (2),
106–124.
Giuliano, G., Narayan, D., 2003. Another look at travel patterns and urban form: the US
and Great Britain. Urban Studies 40 (11), 2295–2312.
Holguin-Veras, J., L´
opez-Genao, Y., Salam, A., 2002. Truck-trip generation at container
terminals: results from a nationwide survey. Transportation Research Record 1790
(1), 89–96.
Hou, Y. and A. Moogoor. 7 Spatial analysis of older adults’ travel behaviour in Singapore.
Urban Environments for Healthy Ageing: A Global Perspective. 2019.
Huntsinger, L.F., Rouphail, N.M., Bloomeld, P., 2013. Trip generation models using
cumulative logistic regression. Journal of Urban Planning and Development 139 (3),
176–184.
Cambridge Systematics Inc., Travel Survey Manual, prepared for US DOT and US EPA.
TMIP Program, Washington DC, 1996.
Kabakus
¸, N., Tortum, A., 2019. Comparative analysis of trip generation models according
to household characteristics for developed, developing and non-developed provinces
in Turkey. S¯
adhan¯
a 44 (5), 1–13.
S. Abu–Eisheh et al.
Transportation Research Interdisciplinary Perspectives 24 (2024) 101048
16
Karaeen, M., et al., Fuel quality implications on combustion engines in the West Bank
area of Palestine. 2013.
McGuckin, N. and A. Fucci, Summary of Travel Trends: 2017 National Household Travel
Survey N. FHWA-PL-l8-019, Editor. 2018, Federal Highway Administration:
Washington D.C., USA.
Montgomery, D.C., Runger, G.C., 2018. Applied statistics and probability for engineers.
Wiley Hoboken, NJ.
Hassan Moussa, Development of A Trip Generation Model for Gaza City, in Civil
Engineering Department. 2013, M.Sc. Thesis, Islamic University of Gaza.
Naser, M., Faris, H., 2015. Developing trip generation rates for hospitals in Amman.
Jordan. Journal of Civil Engineering 9 (1).
Palestinian Central Bureau of Statistics (PCBS). Localities in Jericho & Al Aghwar
Governorate by Type of Locality and Population Estimates, 2007-2016. 2016 [cited
2023 August 15, 2023]; Available from: http://www.pcbs.gov.ps/Portals/_
Rainbow/Documents/jerich.htm.
Palestinian Central Bureau of Statistics (PCBS). Indicators. 2023 [cited 2023 August 15,
2023]; Available from: https://www.pcbs.gov.ps/site/lang__en/881/default.aspx?
lang=en.
Pani, A., et al., 2018. Modelling urban freight generation: a case study of seven cities in
Kerala. India. Transport Policy 69, 49–64.
Priyanto, S., Friandi, E., 2010. Analyzing of public transport trip generation in
developing countries; a case study in Yogyakarta, Indonesia. International Journal of
Civil and Environmental Engineering 4 (6), 167–171.
Sachs, L., Applied statistics: a handbook of techniques. 2012: Springer Science & Business
Media.
Sahu, P.K., Pani, A., 2020. Freight generation and geographical effects: modelling freight
needs of establishments in developing economies and analyzing their geographical
disparities. Transportation 47 (6), 2873–2902.
Sarsam, S., Al-Hassani, S., 2011. Modeling Household Trip Generation for selected Zones
at Al-Karkh Side of Baghdad City. Journal of Engineering 17 (6), 1462–1472.
Shaaban, K., Kim, I., 2016. Assessment of the taxi service in Doha. Transportation
Research Part a: Policy and Practice 88, 223–235.
Soa, G.G., Ali, A.H.A., Al-Zubaidy, H.A.N., 2012. Trip Generation Modeling for Selected
Zone in Al-Diwaniyah City. Journal of Engineering and Development 16 (4).
The European Investment Bank and the Ministry of Transportation, Systematica,
Politecnica, Idrotec, and Air Support, Road and Transportation Master Plan-West
Bank and Gaza Strip: Draft Final Report. June 2016: Ramallah, Palestine.
U.S. Bureau of Public Roads, Guidelines for Trip Generation Analysis. U.S. Government
Printing Ofce, Editor. June 1967: Washington D.C.
UAE Department of Transport, Trip Generation and Parking Rates Manual. 2015: Abu
Dhabi, United Arab Emirates.
Unviersal Group for Engineering and Consulting, 2012. Analytical Study on Population
Projection, Palestine 2015 and 2050: Final Report. National Spatial Plan, Ramallah,
Palestine.
Veerappan, M., et al., 2020. Analysing and modelling the relationship between air freight
movement and airport characteristics in India. Transportation Research Procedia 48,
74–92.
WSP, Does Car Ownership Increase Car Use? A Study of the Use of Car Parking Within
Residential Developments in London. 2014, Commissioned by the Berkeley Group
United Kingdom.
S. Abu–Eisheh et al.
