An-Najah National University  

Faculty of Graduate Studies 

 

 

 

 

 

Developing Trip Generation Models 

Utilizing Linear Regression Analysis: 

Jericho City as a Case Study 

 

 

By 

Alaa Mohammad Yousef Dodeen 

 

 

 

Supervisor 

Prof. Sameer Abu-Eisheh 

 

 

 

This Thesis is Submitted in Partial Fulfillment of the Requirements for 

the Degree of Master of Roads and Transportation Engineering, 

Faculty of Graduate Studies, An-Najah National University, Nablus, 

Palestine. 

2014 





III 

Dedication 

This research effort is dedicated to my family, friends, and instructors. 

Without their love and support, I could not have achieved this goal. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



IV 

Acknowledgment 

First of all, thanks God! 

My appreciation and thanks are extended to my instructors at An-Najah 

National University. 

My special thanks to Professor Sameer Abu Eisheh for his continuous help, 

support and time in this thesis. 

I would also like to thank the defense committee members for their 

valuable discussions. 

Finally, I thank my friends at Jericho Municipality, who provided me with 

the required data (Arial Photo, Master Plan, and Auto CAD Maps) and my 

friends who helped me in data collection and analysis. 

 

 

 

  



V 

 اإلقرار

م الرسالة التي تحمل العنوان:أدناه مقدالموقعة أنا   

Developing Trip Generation Models Utilizing Linear 

Regression Analysis: 

Jericho City as a Case Study 
 

أقر بأن ما اشتممت عميو ىذه الرسالة ، إنما ىي نتاج جيدي الخاص ، باستثناء ما تمت اإلشارة 
الرسالة ككل ، و أي جزء منيا لم يقدم من قبل لنيل درجة عممية أو إليو حيثما ورد ، و أن ىذه 

 بحث عممي لدى أي مؤسسة تعميمية أو بحثية أخرى.

Declaration 

The work provided in this thesis, unless otherwise referenced, is the 

researcher's own work, and has not been submitted elsewhere for any 

other degree or qualification.  

 

Student’s name:                                                                                :اسم الطالب 

Signature:                                                                              :التوقيع           

Date:                                                                                                     :التاريخ 

 

 

 

 



VI 

Table of Contents 
No. Content Page 

 Dedication III 

 Acknowledgment  IV 

 Declaration V 

 Table of Contents VI 

 List of Tables IX 

 List of Figures XI 

 List of Appendices XII 

 Abstract XIII 

 Chapter One: Introduction 1 

1.1 General Background  2 

1.2 The Problem of Study 8 

1.3 Objectives of the Study 9 

1.4 Study Area: Jericho City 9 

1.5 Thesis Outline  11 

 Chapter Two: Literature Review   12 

2.1 Overview of Trip Generation 13 

2.2 Literature Review of Trip Generation Variables 13 

2.2.1 In the developed Countries 13 

2.2.2 In the developing Countries 22 

2.3 Summary of  Literature Review 27 

 
Chapter Three: Analytical Framework and 

Methodology 
28 

3.1 General Steps of the Methodology 29 

3.2 Methods of Survey 30 

3.2.1 Personal Interview Surveys 31 

3.2.2 Telephone Interviews 32 

3.2.3 Mail-Back Surveys 33 

3.2.4 Online Surveys 33 

3.2.5 Data Collection Method Used 34 

3.3 Sample Size Calculation Methods 34 

3.3.1 Standards of Bureau of Public Roads (BPR) 35 

3.3.2 Sample Size Statistical Formulas 35 

3.4 Overview of Linear Regression Method 36 

3.4.1 The Linear Regression Analysis Process  37 

3.4.2 Regression Model Building Approaches 38 

3.5 Unit of Analysis 39 

3.6 Data Analysis Software 39 

3.7 Model Specification 40 



VII 

3.8 Models Estimation 40 

3.8.1 General Trip Generation Model 41 

3.8.2 Trip Generation Models by Trip Purpose 41 

3.8.3 Temporal Trip Generation Models 41 

3.9 Statistical Tests 42 

3.9.1 Correlation Matrix and VIF: Testing for Multicollinearity 42 

3.9.2 R-Squared: Goodness of Fit 43 

3.9.3 F-Test: Testing Overall Significance of Model 44 

3.9.4 T-Test: Testing Individual Coefficients 44 

3.10 Logical Aspects Used in Model Selection 45 

3.11 Determination of Study Area 45 

3.12 Zoning System 47 

 Chapter Four: Field Survey and Data Collection 54 

4.1 Population of Study Area 55 

4.2 Sample Size of Study Area 56 

4.3 Sampling Method 58 

4.4 Questionnaire Design  58 

4.5 Required Information 59 

4.7 Conducting Field Survey 63 

 Chapter Five: Data Analysis and Results 64 

5.1 Descriptive Data 65 

5.1.1 Descriptive Data of Dependent Variables 65 

5.1.2 Descriptive Data of Explanatory Variables 68 

5.2 General Trip Generation Model 79 

5.2.1 Interpretation of Regression Coefficients 80 

5.2.2 Testing Individual Coefficients: T-Test 81 

5.2.3 
Testing for Multicollinearity: Correlation Matrix and 

(VIF) 
83 

5.2.4 Testing Goodness of  Fit: R-Squared (R
2
) 83 

5.2.5 Testing Overall Significance of Model: F-Test 84 

5.2.6 Model Verification 85 

5.3 Trip Generation Models by Purpose 87 

5.3.1 Work Trip Generation Model 87 

5.3.2 Education Trip Generation Model 88 

5.3.3 Shopping Trip Generation Model 90 

5.3.4 Social Trip Generation Model 91 

5.3.5 Recreational Trip Generation Model 93 

5.4 Temporal Trip Generation Models 95 

5.4.1 Trip Generation Model for Trips Made before 8 AM 95 

5.4.2 Trip Generation Model for Trips Made between 8-9 AM 97 

5.4.3 Trip Generation Model for Trips Made between 9-12 AM 100 



VIII 

5.4.4 
Trip Generation Model for Trips Made between 12 AM - 

4 PM 
102 

5.4.5 Trip Generation Model for Trips Made after 4 PM 104 

 Chapter Six: Conclusions and Recommendations 107 

6.1 Summary and Conclusions 108 

6.2 Recommendations  111 

 References 113 

 Appendices 119 

 ب الملخص 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



IX 

List of Tables 

No. Table Title Page 

Table 3.1 
Standards of Bureau of Public Roads (BPR)for Sample 

Size Calculation 
35 

Table 3.2 ANOVA Test Results Table 44 

Table 3.3 Land Uses and Areas for Traffic Zones 50 

Table 4.1 Localities in the Study and Estimates of Population 56 

Table 4.2 
Number of Households per Traffic Zone for Study Area 

and Sample Size Required 
57 

Table 4.3 
Sample Size Calculation According to Statistical 

Formulas 
57 

Table 4.4 Explanatory Variables Used in the Models 62 

Table 4.5 Dependent Variables Used in the Models 62 

Table 5.1 Descriptive Data for the Total Daily Household Trips 65 

Table 5.2 
Descriptive Data for the Daily Household Trips by 

Purpose 
66 

Table 5.3 
Descriptive Data for the Daily Household Trips by 

Time 
66 

Table 5.4 Distribution of Daily Household Trips by Purpose 66 

Table 5.5 Temporal Distribution of Daily Household Trips 67 

Table 5.6 Descriptive Data for the Household Size 68 

Table 5.7 Gender Distribution for the Sample 69 

Table 5.8 Descriptive Data for the Gender Variable 71 

Table 5.9 Descriptive Data for the Number of Employed Persons 72 

Table 5.10 
Descriptive Data for the Number of Persons Continuing 

Education 
72 

Table 5.11 Distribution of Survey Respondents by Age Groups 73 

Table 5.12 Descriptive Data for the Number of Licensed Drivers 74 

Table 5.13 Distribution of Transportation Vehicles 75 

Table 5.14 
Descriptive Data for the Number of Vehicles Owned 

per Household 
76 

Table 5.15 Descriptive Data for the Monthly Household Income 78 

Table 5.16 
Regression Results for the General Trip Generation 

Model 
80 

Table 5.17 ANOVA Table for the General Trip Generation Model 84 

Table 5.18 General Trip Generation Model Verification 86 

Table 5.19 
Regression Results for the Work Trip Generation 

Model 
87 

Table 5.20 
Regression Results for the Education Trip Generation 

Model 
89 

Table 5.21 Regression Results for the Shopping Trip Generation 90 



X 

Model 

Table 5.22 
Regression Results for the Social Trip Generation 

Model 
92 

Table 5.23 
Regression Results for the Recreational Trip 

Generation Model 
93 

Table 5.24 

Regression Results for the Trip Generation Model 

(Number of Daily Trips Made before 8AM per 

Household) 

96 

Table 5.25 

Regression Results for the Trip Generation Model 

(Number of Daily Trips Made between 8-9 AM per 

Household) 

98 

Table 5.26 

Regression Results for the Trip Generation Model 

(Number of Daily Trips Made between 9-12 AM per 

Household) 

100 

Table 5.27 

Regression Results for the Trip Generation Model 

(Number of Daily Trips Made between 12 AM - 4 PM 

per Household) 

103 

Table 5.28 

Regression Results for the Trip Generation Model 

(Number of Daily Trips Made after 4 PM per 

Household) 

105 

 

  



XI 

List of Figures 
Figure No. Figure Title Page 

Figure 1.1 Sequences of Activities in Transportation Analysis 3 

Figure 1.2 
Origins and Destinations for Traffic Zones in an Urban 

Area 
4 

Figure 1.3 Types of Trips according to Movement in Study Area 5 

Figure 1.4 Position of Jericho City in the West Bank 10 

Figure 3.1 Draft Master Plan of Jericho city (2013) 46 

Figure 3.2 
Major Roads and Natural Barriers in Jericho City Used 

for Specifying Zones Boundaries 
49 

Figure 3.3 Explanation of Commuter Shed 50 

Figure 3.4 Land Uses Plan with the Boundaries of Traffic Zones 51 

Figure 3.5 TAZ's of Jericho City 53 

Figure 5.1 Distribution of Daily Household Trips by Purpose 67 

Figure 5.2 Temporal Distribution of Daily Household Trips 68 

Figure 5.3 Daily Household Trips and Household Size 69 

Figure 5.4 Gender Distribution for the Selected Sample 70 

Figure 5.5 Daily household Trips and Number of Males 71 

Figure 5.6 Daily Household Trips and Number of Females 71 

Figure 5.7 
Daily Household Trips and Number of Employed 

Persons 
72 

Figure 5.8 
Daily Household Trips  and Number of Persons 

Continuing Education 
73 

Figure 5.9 Distribution of Survey Respondents by Age Groups 74 

Figure 5.10 
Daily Household Trips and Number of Licensed 

Drivers 
75 

Figure 5.11 Distribution of Transportation Vehicles 76 

Figure 5.12 
Daily Household Trips and Number of Cars per 

Household 
77 

Figure 5.13 
Daily Household Trips and Number of Bicycles per 

Household 
77 

Figure 5.14 
Daily Household Trips and Number of Motorcycles per 

Household 
78 

Figure 5.15 
Number of Daily Household Trips and Monthly 

Household Income 
79 

 

 

 

 

 



XII 

List of Appendices 
Appendix Appendix Title Page 

Appendix A Correlation Matrix  120 

Appendix B Questionnaire Form  121 

Appendix C SPSS Results (General Trip Generation Model) 125 

Appendix D SPSS Results (Work Trip Generation Model) 126 

Appendix E SPSS Results (Education Trip Generation Model) 127 

Appendix F SPSS Results (Shopping Trip Generation Model) 128 

Appendix G SPSS Results (Social Trip Generation Model) 129 

Appendix H SPSS Results (Recreational Trip Generation Model) 130 

Appendix I 
SPSS Results  (Trip Generation Model for Trips Made 

before 8 AM) 
131 

Appendix J 
SPSS Results (Trip Generation Model for Trips Made 

between 8-9 AM) 
132 

Appendix K 
SPSS Results (Trip Generation Model for Trips Made 

between 9 AM - 12 PM) 
133 

Appendix L 
SPSS Results (Trip Generation Model for Trips Made 

between 12 - 4 PM) 
134 

Appendix 

M 

SPSS Results (Trip Generation Model for Trips Made 

after 4 PM) 
135 

 

 

 

 

 

 

 

 

 

 

 

 



XIII 

Developing Trip Generation Models Utilizing Linear Regression 

Analysis: 

Jericho City as a Case Study 

By 

Alaa Mohammad Yousef Dodeen 

Supervisor 

Prof. Sameer Abu-Eisheh 

 

Abstract 

The aim of this research is to develop trip generation models to predict the 

number of trips generated by households in the Palestinian areas 

considering Jericho City as the case study. The models are developed using 

multiple linear regression analysis, which establishes relationship between 

the number of trips generated by households and some socioeconomic 

attributes. 

The developed models include three types of models. The first model is a 

general trip generation model (i.e., a general model regardless of trip 

purpose and trip time). The second one includes trip generation models by 

trip purpose. These models include the work trip generation model, the 

education trip generation model, the shopping trip generation model, the 

social trip generation model, and the recreational trip generation model. 

Finally, five trip generation models by trip time are developed.   

The data consists of primary data, which was collected by conducting a 

household survey. The survey consists of 713 randomly selected 

households from Jericho City, the study area. 

The results indicated that the estimated general trip generation model has a 

good explanatory power. The R-square for this model is 0.69, indicating 



XIV 

that the explanatory variables included in the model explain 69% of the 

dependent variable. The variables that mostly affect trip generation are 

found to be the number of persons receiving education in the household, 

the number of employed persons in the household, as well as the household 

monthly income. 

The work trip generation model has R-square value of 0.74. In this model, 

the number of employed persons in the household and the number of 

persons with age between 31 to 50 years are the variables that mostly affect 

work trips. The educational trip generation model has R-square value of 

0.97. The number of persons who are receiving education in the household 

is the main factor in this model.  

The shopping trip generation model depends on the number of persons in 

the household and the monthly household income. The social trip 

generation model depends mainly on the number of females in the 

household and the number of employed persons in the household. Finally, 

the recreational trip generation model depends mainly on the number of 

persons receiving education in the household, number of persons between 

51 and 64 years old, and the monthly household income. 



1 

 

 

 

 

 

 

 

 

Chapter One 

Introduction 

 

 

 

 

 

 

 

 



2 

Chapter One 

Introduction 

1.1 General Background  

Transportation planning processes have been intensively used to estimate 

the demand for travel encountered in the future. The estimated travel 

demand is utilized as a basis to plan for future transportation facilities and 

services. 

As for the transportation system, it is necessary to quantify the inputs and 

the outputs for the system. The system inputs are the quantum of demand 

for transportation in the future years, while the system outputs are system 

characteristics that are planned for meeting the demand on the horizon 

years (Hutchinson, 1974). 

In order to quantify the inputs and outputs, a major and commonly used 

planning method suggests that there are four analytical steps followed to 

get the total demand in the horizon year, which is called Urban Traffic 

Management System (UTMS). The characteristics of the system that is 

proposed for the horizon year condition are to be assumed. The analytical 

steps are: trip generation, trip distribution, modal split, and route 

assignment. These steps are shown in Figure 1.1. 

This research considers the first step of this sequence, which is related to 

trip generation where it will be thoroughly studied and analyzed for Jericho 

City as a case study of Palestinian cities. The expected outputs can be 

applied to other Palestinian cities with proper calibrations. However, how 

the model can be applied to other cities is out of the scope of this research.  



3 

 

Figure 1.1: Sequences of Activities in Transportation Analysis 

Source: Principles of Urban Transport Systems Planning (Hutchinson, 1974) 

Trip generation analysis means understanding the trips generated in 

different traffic zones in an urban area. The trip for the purpose of analysis 

is defined as a one-way movement from an origin to a destination for a 

person. 

The entire urban area is usually divided into smaller traffic zones, where 

the points of origins and destinations are fixed as zone centroid. This is 

illustrated in Figure1.2. 



4 

 

Figure 1.2: Origins and Destinations for Traffic Zones in an Urban Area 

After delineating the urban area boundary and fixing the zone centroids as 

points of origin or points of destination, the trips can be classified 

according to spatial movement into four kinds (Meyer and Miller, 1986): 

1. Internal to internal trip 

2. Internal to external trip 

3. External to internal trip 

4. External to external trip 

These types of trips according to movement are shown in Figure1.3. 

Trip generation analysis is mainly related to the internal to internal trips 

and less to internal to external trips. Home interview survey is the major 

tool to establish trip generation models as will be discussed later. The other 

kinds (internal to external, external to internal, and external to external) can 

be identified by cordon surveys, which means surveys conducted at 

convenient points on the cordon line or at points of intersection of radiating 

roads.   



5 

 

Figure 1.3: Types of Trips according to Movement in Study Area 

The principal task in trip generation analysis is to relate the intensity of trip 

making (number of trips made from a point to several other points) to and 

from traffic zones to measures of the type and intensity of the land use in 

these zones and to other socio-economic characteristics.   

There are two types of trip generation analysis that can be carried out; 

these are: 

1. Trip production analysis 

2. Trip attraction analysis 

The term trip production refers to the trips generated by residential zones, 

where these trips are either trip origins or trip destinations (Papacostas and 

Prevedouros, 2004). Trips, which end at home, are called home-based trips 

or trip production, while the term trip attraction is used to describe trips 

generated by activities at the non-home ends (Hutchinson, 1974). 

Home-based trips and non-home based trips are analyzed separately, 

because it is difficult to combine these categories of trips in developing 

models. There is a need to develop separate trip generation models for 



6 

home and non-home based trips, as the type of variables that might 

influence such trips are to be different, or if the variables are the same, the 

effect of these variables on trip making might be different, so the models 

for these two types must be separated (Papacostas and Prevedouros, 2004). 

The process of relating the trips produced by households to the factors 

influencing trip production by appropriate analytical technique is termed as 

trip production modeling and the process relating the trips attracted to non-

residential ends to the factors influencing trip attraction by appropriate 

analytical technique is termed as trip attraction modeling (Papacostas and 

Prevedouros, 2004). 

Although the individual is usually the trip makers, numbers of trips per 

household are usually estimated. The household is defined according to the 

Palestinian Central Bureau of Statistics (PCBS), as one person or a group 

of persons with or without family relationship, who live in the same 

housing unit, share meals, and make joint provision of food and other 

essentials of living. This is the general condition, where the persons may be 

related or unrelated to one another or both. The unrelated persons are called 

institutional household and are also considered in this study (Palestinian 

Central Bureau of Statistics, 2003). 

Some factors that usually influence trip production are related to the 

household characteristics because the household is a major unit of trip 

production. These can be listed as follows (Arasan, 2012): 



7 

1. Household size and composition, where household size is number of 

persons in household, and composition like the average age of 

household, the distribution of the sex of the individuals 

2. Number of employed persons 

3. Number of students 

4. Household income 

5. Vehicle ownership 

6. Number of persons in household who have driving license 

7. Type of house if it is independent or apartment 

The factors that usually influence trip attraction are related to non-home 

zones or non-residential zones like commercial zones, industrial zones, 

institutional zones, and recreational zones. These factors, which reflect the 

type and the intensity of land-use, can be as follows (Arasan, 2012): 

1. Retail trade floor area  

2. Service and office floor area  

3. Manufacturing and wholesales floor area  

4. Number of employment opportunities in retail trade 

5. Number of employment opportunities in service and offices  

6. Number of employment opportunities in manufacturing and whole 

sale 

7. School and college enrollment 



8 

8. Number of special activity centers like transport terminals, sport 

stadium, major recreational, cultural, and religious places 

On the other side, trip production or home-based trips can be classified into 

different categories based on trip purpose, which are: 

1. Work trips 

2. Education trips 

3. Shopping trips   

4. Social and recreational trips  

Finally, trip production or home-based trips can also be classified into 

different categories based on time of making the trip. 

According to the level of analysis, it could be analyzed on an aggregate 

level (zone or area), or disaggregate level (household or person). In this 

research, analysis on the level of households is adopted, because depending 

on this level, more accurate results can be obtained, through the study of 

movements for each household. 

1.2 The Problem of Study 

There is a lack of specialized studies, which are related to quantifying and 

modeling travel demand in the Palestinian cities.  

Several factors led to considerable increases in trip generation in the 

Palestinian cities, including reduced taxes on prices of cars, which have led 

to the increase in their numbers, as well as the increasing need for mobility 

according to population growth. This is combined with the limited 

development of the transport networks within the cities, and the observed 

overloading of and congestion in the existing transport networks.  



9 

As the numbers of cars are expected to continue increasing in the future, 

the need of transportation planning appears to be essential for these cities. 

Proper modeling is lacking describing the four analytical steps; trip 

generation, trip distribution, modal split, and route assignment.  

To provide strong basis for the transportation planning process, this 

research considers the need for studies in this area, through studying the 

trip production part of trip generation stage and its various categories. 

1.3 Objectives of the Study  

The aim of this study is to lay the basis to predict current and future traffic 

trips generated from different traffic zones that comprise a Palestinian city, 

thus studying and modeling trips produced from households according to 

their characteristics, relying on the principles of the regression analysis 

technique. It also aims to predict the number of trips according to trip 

purpose and trip timing of trip production or home-based trips for different 

traffic zones. 

1.4 Study Area: Jericho City 

Jericho City is chosen to be the study area. It is located in the middle part 

of the West Bank. The city is considered medium-sized in area. It has a 

total area of 57.43 km
2
 (Palestinian Central Bureau of Statistics, 2012). 

According to the latest population estimate conducted by the Palestinian 

Central Bureau of Statistics (PCBS) in 2012, the population of Jericho City 

was estimated to be about 20,253 people. The number of households was 

estimated to be 3,510 living in 3,386 buildings. The population density in 

Jericho City was estimated to be 1, 055 inhabitants/km
2
. In the refugee 



10 

camps- in Jericho, the population density ranges between 3,763-4,190 

inhabitants/km
2
. According to these statistics, Jericho City needs to expand 

between 2,000 and 7,000 acres. 

According to the PCBS statistics, there are nearly 3,386 buildings in 

Jericho City containing 4,549 units. Around 84.4% of these buildings are 

the property of their owners. The city includes the city center, the old 

residential area, areas of moderate density expansion, urban areas, newly 

developed suburbs, as well as refugee camps. Each of these styles is 

different from the other depending on the purpose, number of housing 

units, the nature of the buildings, and population density. There are many 

housing projects in progress in the city. The position of Jericho City in the 

West Bank is shown in Figure 1.4. 
 

 

 

 

 

 

 

 

 

 

 

 

Figure 1.4: Position of Jericho City in the West BankSource: The Palestinian Central 

Bureau of Statistics (2010) 



11 

1.5 Thesis Outline  

This thesis is divided into six chapters. Chapter one presents a general 

background, the problem of the study, the objectives of the study, study 

area, and thesis outline. Chapter Two is concerned with the literature 

review. Chapter Three discusses the methodology, while Chapter Four 

describes the field survey and data collection. Chapter Five addresses data 

analysis and results. Finally, Chapter Six presents the main conclusions and 

recommendations of the study. 

 

 

 

 

 

 

 

 

 

 

 

 



12 

 

 

 

 

 

 

 

 

Chapter Two 

Literature Review 

 

 

 

 

 

 

 

 



13 

Chapter Two 

Literature Review 

2.1 Overview of Trip Generation 

Trip generation analysis involves estimation of the total number of trips 

entering or leaving a parcel of land as a function of the socio-economic, 

locational, and land use characteristics of the parcel. The function of trip 

generation analysis is to establish meaningful relationships between land 

use and trip making activity so that changes in land use can be used to 

predict subsequent changes in transport demand (Paquette and Ashford, 

1982). 

2.2 Literature Review of Trip Generation Variables 

This section addresses the literature of the explanatory variables that are 

included in trip generation models in studies conducted both in developed 

as well as developing countries. 

Indeed, several research papers studied the variables that affect trip 

generation and found some significant relationships between trip 

generation and the above mentioned characteristics. Most of the empirical 

research has been conducted in the US and other developed countries. 

However, some empirical research has been conducted in other developing 

countries. 

2.2.1 In the Developed Countries 

Hunt and Broadstock (2010) constructed a trip generation model for a cross 

section of residential developments around the UK. The empirical model 

tested whether trip making patterns for residential developments are 

http://eprints.port.ac.uk/view/local_creators/5162.html
http://eprints.port.ac.uk/view/local_creators/5072.html


14 

independent of car ownership. The result was that trip generation is 

dependent upon car ownership, socio-economic factors and site-specific 

characteristics, in particular land-zone type. However, public transport 

services are not found to have a significant relationship with trip 

generation. Consequently, a policy implication of the results was that 

increasing bus services to residential developments is not associated with a 

reduction in generated trips. 

Guiliano (2003), Guiliano and Dargay (2006), and Guiliano and Narayan 

(2003) found significant differences in travel behavior between different 

demographic groups in the USA and the UK. Their data showed that 

American participants made 4.4 trips per day travelling approximately 31 

miles (49.6 km), whereas the British participants travelled only 16 miles 

(25.6 km) in 3 trips per day. The method of data collection however varied 

between the US and the British studies. The USA data was collected by 

telephone using a stratified sample with participants using a one day 

„recall‟ diary. British participants were selected using a stratified random 

sample based on post code. The British participants were interviewed 

directly and were required to complete a seven day travel diary. Thus, 

methods of sampling and data collection varied between the two groups.  

The authors found that participants aged 65 years or older in the UK 

travelled half the distance and were less likely to travel on any given day 

than participants aged 18 – 64 years. In the USA study, participants aged 

65 years or more travelled 60% of the distance of the younger participants. 

The authors also suggested that lower household incomes in the UK 



15 

compared with the USA produced lower travel demand and car ownership. 

Gender, age, and household income were all found to influence travel 

behaviors and that there were significant differences between travel 

behaviors in the USA and the UK. It was found that the significantly higher 

transport costs in the UK led to a decrease in trips. 

Newbold et al. (2005) studied the travel behaviors of Canadians aged 65 

years or more to determine if their travel patterns were different from 

younger Canadians. Their study used data from the General Social Survey 

(GSS) of Canada. The data from approximately 19,000 participants 

provided a partial confirmation of the research question but recognized that 

factors other than age can influence travel behavior.  

The results showed that older Canadians make fewer daily trips than 

younger Canadians, but as expected, this could be caused by the fact that 

the participants in the study were no longer employed and hence were no 

longer making travel-to-work journeys. Thus, daily trip numbers and 

duration decreased significantly due to changes in employment and health 

status. In addition, there was a greater reliance on the car and a significant 

reduction in the use of public transport as the principal travel mode 

compared with younger Canadians. 

Best and Lanzendorf (2005) conducted an empirical study in Cologne, 

Germany to determine if there were gender differences in car use and travel 

patterns for maintenance travel.  

Overall, the authors found that there were no significant differences in the 

total number of trips or distances travelled between men and women. 



16 

However, the type or destination of trips did provide some gender 

differences.  

The authors found that women made fewer journeys to work by car and 

more journeys for non-work activities such as shopping and child-care. 

This was also confirmed by Boarnet and Sarmiento (1998) in their study of 

travel behavior in southern California. 

Moriarty and Honnery (2005) studied urban travel in all Australian State 

capital cities. Although the major emphasis was on studying the 

relationship between the distance from place of residence to the Central 

Business District (CBD) of each city and the impact on travel behavior. 

Their study found that women on average travel less often and for shorter 

distances than men. 

Olaru et al. (2005) studied the travel behavior in the Sydney Metropolitan 

area, Australia and found that a number of socio-demographic variables 

influenced travel behavior. Women were more likely to travel closer to 

home than men, particularly if they came from a non-English speaking 

household. 

Simma and Axhausen (2004) conducted a study to explore the impacts of 

personal characteristics and the spatial structure on travel behavior, 

especially mode choice in Upper Austria. The spatial structure is described 

among other things by accessibility measures. The models were estimated 

using structural equation modeling (SEM). The models were based on the 

1992 Upper Austrian Travel Survey and the Upper Austrian Transport 

Model. 



17 

The results highlighted the key roles of car ownership, gender and work 

status in explaining the observed level and intensity of travel. The most 

important spatial variable was the number of facilities, which can be 

reached by a household. The municipality-based variables and the 

accessibility measures have rather little explanatory power. 

The reasons for this low explanatory power were considered. Although the 

findings in this study indicated that the spatial structure is not a decisive 

determinant of traffic. The results provided useful hints for possible policy 

alternatives. 

Polk (2004) wrote a paper to test the influence of gender on daily car use 

and on willingness to reduce car use in Sweden. Car use was modeled in 

terms of practical factors combined with manifestations of the specific 

influence of gender. Willingness to reduce car use was modeled in terms of 

attitudinal factors using a theory of environmentalism.  

The results confirmed the existence of a gender component. The author 

found a significant relationship between sustainable travel patterns and 

gender. Women were more willing to reduce their use of the car than men, 

more positive towards reducing the environmental impact of travel modes, 

and more positive towards ecological issues. 

The concluding discussion suggested that more research is needed to 

further theoretical understanding and methodological expertise regarding 

how gender can be modeled in travel research in order to attain current 

policy regarding gender equal transportation system. 



18 

Georggi and Pendyala (2001) conducted a detailed analysis of long-

distance travel behavior in the USA for two key socio-economic groups of 

the population: the elderly and the low income. The analysis utilized data 

from the 1995 American Travel Survey that provided a rich source of 

information on long-distance travel undertaken over a period of 12 months.  

The analysis focused on comparing the elderly and the low-income groups 

of the population against other groups with respect to various demographic 

and trips characteristics. The travel behavior comparison included an 

analysis by trip purpose, travel mode, distance, trip duration, and trip 

frequency. In addition, regression models of long-distance trip generation 

were estimated separately for different groups to examine differences in 

trip generation propensity across the groups. 

The results showed that both the elderly and the low income undertake 

significantly fewer long-distance trips than other socio-economic groups. It 

was found that nearly half of the low income and elderly made no long-

distance trips in the one-year survey period. In addition, it was found that 

long-distance trips made by these groups were more likely to be 

undertaken by bus and geared towards social and personal business 

activities.  

Badoe and Steuart (1997) estimated the total number of household 

shopping trips for the Greater Toronto Metropolitan area, Canada by using 

such variables as household size, number of workers, number of licensed 

persons, and number of vehicles. Regression analysis was used with 1964 

and 1986 data for the Greater Toronto Metropolitan area.  



19 

The different model specifications provided little explanation of the 

variation in household shopping trips. The household size was found to 

have little explanatory power. The authors concluded that different 

approaches are needed to explain the variation in non-work trips, including 

shopping trips. 

Golob (1989) undertook a study in Germany to model the causal 

relationship, at the household level, among income, car ownership, with 

trip generation. The study was based on data obtained from the Dutch 

National Mobility Panel, which consisted of approximately 1,800 

households, stratified by life cycle. 

The author found that car ownership directly affects public transport trip 

making, with additional effects from income. The author also stated that 

there are direct links from the lowest and highest income categories to 

public transport demand. 

Vickerman and Barmby (1984) analyzed the relationship between shopping 

expenditure and shopping trips. The data used in their study were collected 

as part of an earlier research project on shopping travel. The study sample 

consisted of 1,074 households in 25 districts in the County of Sussex, 

England. 

The authors estimated the number of weekly shopping trips and the total 

weekly shopping expenditures using simultaneous equations. The 

explanatory variables included household size, income, and auto 

ownership. It was found that income has little effect on trip making, but 

that accessibility and the cost of travel are important factors.  



20 

Stopher and McDonald (1983) conducted a trip generation analysis on data 

from the Midwest, USA using multiple classification analysis (MCA) in 

contrast to linear regression analysis. The household-structure variable was 

tested using both analysis of variance and MCA to determine how well the 

variable performs in various model structures when compared with other 

variables. The variables tested were the number of cars or vehicles 

available to the household, household size, housing type, total number of 

employed persons, household income, and total number of licensed drivers. 

The analysis concluded that the household-structure variable did not 

perform significantly better than the other variables tested. 

Supernak et al. (1983) presented a person-category model of trip 

generation as an alternative to household-based trip-generation models in 

Washington, D.C., USA. In their model, a homogeneous group of persons 

was used as an analysis unit. The variables of age, employment status, and 

automobile availability were found to be the most significant descriptors of 

a person's mobility. The final version of the model was based on eight 

person categories.  

Oldfield (1981) carried out an analysis on the effect of income on bus 

travel in the UK. The study was based on Family Expenditure Surveys and 

the National Travel Survey and considered mainly the non-car-owning 

households. The author examined some cross-sectional data on bus travel 

as a function of household income, and the way in which it changed over 

time, with the intention of seeing whether demand elasticity with respect to 

earnings can be estimated. Unfortunately, no consistent pattern had 



21 

emerged from the analysis to enable demand elasticity with respect to 

earnings to be fixed with precision. 

Downes et al. (1978) conducted a study on household and person trip 

generation model. The authors used a single data bank, comprising over 

60,000 trips from sample household surveys in the UK to compare two 

alternative types of trip generation models; one based on household trip 

rates and the other on person trip rates for each household. Their 

performance was found to be similar, each accounting for over 50% of the 

variability in household trip rates, but the person trip rate model had been 

shown to be simpler to use and statistically more acceptable.  

The most important variables for modeling home-based trips were 

household size and car ownership in both types of models. Work trips 

required only household employment in a household rate model and car 

ownership in an employed person rate model. Household location and the 

year of study had a small but discernible effect on trip rates due to some 

reduction in the inner and middle area rates between the two years.  

Robinson and Vickerman (1976) were concerned with the development of 

a basic methodology for the future study of shopping. The authors 

demonstrated how the existing methodology failed to allow for some of the 

most important dimensions of the shopping choice decision. A revised 

model was developed and tested against especially collected data from a 

cross-section of households in Sussex, South East England. The authors 

demonstrated the importance of attraction and accessibility in determining 

variations in levels of household shopping activity. 



22 

2.2.2 In the Developing Countries  

Moussa (2013) conducted a study to develop a trip generation model for 

Gaza City, Palestine to determine the household travel characteristics 

pattern in the study area. The study also aimed to compare trip rates as 

modeled by the Conventional Cross Classification (CCA) method with that 

of Multiple Cross Classification (MCA) method in Gaza City. 

Furthermore, the researcher aimed to develop a trip attraction model using 

Multiple Linear Regression technique (MLR).  

Household interview survey was conducted to collect the primary data. The 

survey was distributed to 425 households in different districts of Gaza City. 

The results indicated that vehicle ownership, household size, income level, 

and the number of licensed drivers in the household are the main factors 

that affected trip production in Gaza City.  

In addition, the results showed that the Multiple Cross Classification 

(MCA) models are more effective in expressing trip rates for trip 

production than the Conventional Cross Classification (CCA) models. 

Sofia et al. (2012) developed a relationship between the daily household 

trips and socio-economic characteristics for Al-Diwaniyah City, Iraq. The 

authors used the stepwise regression technique (multiple linear regression) 

after the collected data had been fed to the SPSS software. The city was 

divided into 5 sectors with 70 zones covering an area of 52 km
2
.  

Home questionnaire forms were distributed through arrangements with the 

secondary, industrial, commercial school administrations, and some 

colleges. The results showed that the trip generation model mainly depends 



23 

on family size, gender, the number of workers, and the number of students 

in the family. 

Sarsam and Al-Hassani (2011) developed statistical models to predict trip 

volumes for a proper target year in Al-Karkh side of Baghdad City, Iraq. 

Non-motorized trips were considered in the modeling process. The 

traditional method to forecast the trip generation volume according to trip 

rate, based on family type, was used in the study. Families were classified 

by three characteristics: social class, income, and number of vehicle 

ownership. The study area was divided into 10 sectors. Each sector was 

subdivided into a number of zones. The total number of zones was 45 

based on the administrative divisions. The trip rate for the family was 

determined by sampling. A questionnaire was designed and interviews 

were made for data collection from the selected zones. 

Two techniques were used, full interview and home questionnaire. The 

questionnaire forms were distributed in many educational institutes 

including intermediate, secondary, and commercial schools. The developed 

models were total trips per household, work trips per household, education 

trips per household, shopping trips per household, and social trips per 

household. These models were developed using stepwise regression 

technique after the collected data had been fed to the SPSS software. 

The results showed that total trips per household are related to the family 

size and the structure variables such as the number of persons who are 

above 6 years of age, the number of males, the total number of workers, the 

total number of students in the household, and the number of private 



24 

vehicles. The model had coefficient of determination equals to 0.67 for the 

whole study area.  

The results also showed that the home-based work trips are related to the 

number of workers in the household, number of male workers in the 

household, number of female workers in the household, and number of 

persons between 25 and 60 years of age. This model has coefficient of 

determination equals to 0.82 for the whole study area. Home-based 

education trips are strongly related to the number of students in the 

household and this model had coefficient of determination equals to 0.90 

for the whole study area. 

Priyanto and Friandi (2010) developed a trip generation model for public 

transport passengers in Yogyakarta, Indonesia, using multiple linear 

regression analysis. The authors established a relationship between the 

number of trips and socio-economic attributes. The data consisted of 

primary and secondary data. Primary data were collected by conducting 

household surveys, which were randomly selected.  

The resulted model showed that public transportation trips seem to have 

negative correlation with income, motorcycle ownership, and car 

ownership. This means that the number of trips made by people decreases 

as income, the number of motorcycles, and cars owned increase. It is 

different from the general trip generation model (the model for all trips by 

either private or public transportation) where the number of trips 

commonly rises with the increase in income, motorcycle, and car 

ownership.  



25 

The model also showed that the number of public transport trips increases 

as the family size increases. Commonly, the higher the number of family 

members, more public transport trips will be made. 

Pettersson and Schmocker (2010) conducted a study to analyze travel 

patterns by those aged 60 or over in Metro Manila, The Philippines. Trip 

frequency and tour complexity were analyzed with ordered probit 

regression, separating the effects of socio-demographic characteristics as 

well as land-use patterns. The results were compared to observations made 

for cities in developed countries, in particular London as an example for a 

city in a first world country.  

The authors showed that there is a pronounced decrease in total trips made 

with increasing age in Manila. However, analyzing for specific trip 

purposes, the authors found, similarly to trends in developed countries, that 

the number of recreational trips is fairly constant in all age groups. 

Said (1990) conducted an empirical study to estimate work trip rates for 

households in Kuwait using a generalized linear model (GLM). Seven 

different household groups were identified from the 1985 census. One of 

these groups, Kuwaiti households living in villas, was used for some 

illustrative GLM analysis. The study showed that work trip rates of this 

household group are influenced by car ownership, household size, and the 

interactive effect of these two variables. 

Deaton (1985) undertook an empirical analysis on income and trip making 

for developing countries and cities including Hong Kong, Sri Lanka, 

Tunisia, Thailand, and several suburbs of Delhi, India. In this paper, the 

http://www.sciencedirect.com/science/article/pii/019126079090056C


26 

author illustrated that data from household surveys from developing 

countries and cities could be used for analyzing the relationship between 

the household income/expenditure and the trip characteristic. The author 

also suggested that households spending anything on travel tended to 

increase with income in all surveys. The author also mentioned that 

conditional on vehicle ownership; both private and public transport 

demands were income elastic. 

Maunder (1981) undertook extensive household surveys in two different 

income communities in Delhi, India. The two study areas were Shakarpur 

and West Patel Nagar. Shakarpur was a low/middle income unauthorized 

residential community while West Patel Nagar consisted of both middle 

and upper income residents. Although the average monthly household 

income in West Patel Nagar was 50% more than that in Shakarpur, the trip 

rate in the two communities was similar. However, it was worth noting that 

both communities had a large proportion of household in possession of 

bicycles (i.e. 30% in West Patel Nagar and 48% in Shakarpur). Also, in 

Shakarpur, the bicycle was the major personal vehicle. 

Heraty (1980), with the cooperation of the Jamaican Government, carried 

out a study on the public transport in Kingston, Jamaica and its relation to 

low income households. In the study, the author analyzed the role public 

transport played in the life style of low-income households from the 

Jamaican Government survey data and in depth research with low-income 

households. The author concluded that household expenditures on travel 

increased with income. However, the percentage of total household income 



27 

spent on travel was greater for low-income families. The author also 

mentioned that car ownership was strongly related to income so that the 

relationship of public transport was greater for the lower income group. 

2.3 Summary of Literature Review  

To summarize, a range of socio-demographic variables influences trip 

generation. These variables include gender, age, number of children, 

household income, employment status, number of workers in the 

household, retirement status, education status, vehicle ownership, family 

size, and type of house. 

In terms of the methods that are used in developing trip generation models, 

the primary tool used in modeling trip generation is the regression analysis 

method. 

Two types of regression are commonly used. The first uses data aggregated 

at the zonal level, with average number of trips per household in the zone 

as the dependent variable and average zonal characteristics as the 

explanatory variables. The second uses disaggregated data at the household 

or individual level, with the number of trips made by a household or 

individual as the dependent variable and the household and personal 

characteristics as the explanatory variables. 
 

 

 

 



28 

 

 

 

 

 

 

 

 

Chapter Three 

Analytical Framework and Methodology 

 

 

 

 

 

 

 

 



29 

Chapter Three 

Analytical Framework and Methodology 

3.1 General Steps of the Methodology 

The research will mainly rely on the following methodology: 

1. Desk and Internet Research: This step is done by reviewing the 

related literature including publications on trip generation analysis 

using different methods including the linear regression method.   

2. Study Area Selection: This involves selecting the boundaries of the 

study area and dividing the city into traffic zones according to set 

criteria and benefiting from maps made available by the 

Municipality of Jericho or other relevant agencies. 

3. Selecting Sample Size and Designing Household Questionnaire: 

This involves identifying needed information and designing a 

questionnaire for proper data collection of relevant socio-economic 

variables and trip data.  

4. Collecting the Required Data: This is related to field data 

collection from the sample of households in the different traffic 

zones. Required data include the total number of trips made by 

household members, number of trips made by household for each 

category of trip production, whether according to trip purpose or 

trip time, through home interview survey.  

5. Analyzing Available Data: This is performed depending on 

relevant computerized programs. 



30 

6. Building Models: This includes estimation of feasible models for 

the total number of trips produced by the household, and the total 

number of trips for each category of home-based trips (either 

according to trip purpose or trip time) utilizing linear regression 

method. 

7. Selection of Appropriate Models: This is done using statistical and 

rational methods available to choose the most appropriate model 

that predicts the total number of trips produced by the household, 

and the models that predict the number of trips for each category of 

home-based trips. 

8. Verification of Results: This is done by comparing the outputs of 

models with the actual numbers to verify these models, and make 

calibration of the variables used in model building. 

The details of steps of methodology are shown in the following sections 

and chapters. 

3.2 Methods of Survey 

Once the questionnaire is ready, the next step is to conduct the actual 

survey. There are various types of survey methods, and each has its own 

advantages and disadvantages. Before choosing the proper survey method, 

there is a need to know the characteristics of the people that will be 

surveyed, the sample size, the cost, the expected response rate, and the size 

of study area. For example, if the target population is composed of college 

students, the online survey method would be preferable. In terms of sample 

size, the number of respondents in the sample should be considered when 



31 

choosing the survey method. Online surveys or mail-back surveys are best 

for surveys requiring a hundred or thousand responses. On the other hand, 

telephone surveys are ideal for studies requiring limited responses. If the 

cost or size of study area is limited, the online or mail-back surveys are 

preferable. For better response rates, the personal interview survey is 

preferable. 

The most common methods of survey are discussed below. 

3.2.1 Personal Interview Surveys 

In this method, also known as a face-to-face survey, the researcher visits 

the home of the respondent and asks the questions and fills up the 

questionnaire by himself. This method is considered to be the most 

effective one. The purpose of conducting a personal interview survey is to 

explore the responses of the people to gather more and deeper information 

(Sincero, 2012). 

There are two types of personal interview surveys according to how the 

interviewer approaches the respondents: intercept and door-to-door 

interviews. In an intercept interview, the interviewer usually conducts a 

short but concise survey by means of getting the sample from public places 

such as malls, theaters, food courts, or tourist spots. On the other hand, a 

door-to-door interview survey involves going directly to the house of the 

respondent and conducting the interview either on-the-spot or at a 

scheduled date (Sincero, 2012). 

The personal interview surveys have many advantages including high 

response rate, more precise data, and better observation of respondents' 

behavior. However, personal interview surveys are more expensive than 



32 

other types of surveys and more time-consuming because there is a need to 

travel and meet the respondents at different locations. 

3.2.2 Telephone Interviews 

Conducting interviews by telephone may offer potential advantages 

associated with access, speed and lower cost. This method allows the 

researcher to make contact with participants with whom it would be 

impractical to conduct an interview on a face-to-face basis because of the 

distance and prohibitive costs involved and time required. Even where 

„long-distance‟ access is not an issue, conducting interviews by telephone 

may still offer advantages associated with speed of data collection and 

lower cost. In other words, this approach may be seen as more convenient.  

However, there are a number of significant issues that are taken into 

account against attempting to collect qualitative data by telephone contact.  

Seeking to conduct qualitative interviews by telephone may lead to issues 

of reduced reliability, where the participants are less willing to engage in 

an exploratory discussion, or even a refusal to take part. There are also 

some other practical issues that would need to be managed. These relate to 

the researcher's ability to control the pace of a telephone interview and to 

record any data that were forthcoming.  

Conducting an interview by telephone and taking notes is an extremely 

difficult process and so using audio-recording is recommended. In 

addition, with telephone interviews the researcher loses the opportunity to 

witness the non-verbal behavior of participants, which may adversely 

affect the interpretation of how far to pursue a particular line of 

http://explorable.com/types-of-survey


33 

questioning. Participants may be less willing to provide the researcher with 

as much time to talk to them in comparison with a face-to-face interview. 

The researcher may also encounter difficulties in developing more complex 

questions in comparison with a face-to-face interview situation.  

For the above mentioned reasons, interviewing by telephone is likely to be 

appropriate only in particular circumstances. It may be appropriate to 

conduct a short, follow-up telephone interview to clarify the meaning of 

some data, where the researcher has already undertaken a face-to-face 

interview with a participant. It may also be appropriate where access would 

otherwise be prohibited because of long distance. 

3.2.3 Mail-Back Surveys 

The researcher sends the questionnaire to the respondents and asks them to 

fill the details and mail them back with required information. Care should 

be taken to design the questionnaire so that it is self-explanatory. 

The mail-back surveys save the researchers' time. In addition, this method 

is considered a quick process of collecting data. It involves minimal cost in 

comparison with other methods of survey like face-to-face surveys. 

The main disadvantages of mail-back surveys are that they need more 

effort in questionnaire designing because in order to be clear and concise to 

encourage every respondent to reply. This method has a low response rate 

and generates less precise data in comparison with other methods of 

survey.  

3.2.4 Online Surveys 

For the past few years, the Internet has been used by many companies in 

conducting all sorts of studies all over the world. Whether it is market 



34 

or scientific research, the online survey has been a faster way of collecting 

data from the respondents as compared to other survey methods such as 

personal interviews (Sincero, 2012). 

Advantages of online surveys include ease of data gathering, minimal cost, 

automation in data input and handling, an increase in response rate, and 

flexibility of questionnaire design. 

However, online surveys have some disadvantages such as the absence of 

interviewer, inability to reach some target groups of population, and survey 

fraud. 

3.2.5 Data Collection Method Used 

The method of survey that is used in this research to collect data is the 

personal interview (door to door). This method is used because it generates 

the highest response rate and the most precise data. In addition, the area of 

study is not so large and the target population is comprised of all kinds of 

people. 

3.3 Sample Size Calculation Methods 

Unfortunately, there is no straightforward and one objective answer to the 

question of the calculation of sample size. Determining the sample size is a 

problem of trade-offs (Ortuzar and Willumsen, 1996): 

1. A much too large sample may imply a data-collection and analysis 

process, which is too expensive given the study objective and its 

required degree of accuracy. 

2. A far too small sample may imply results which are subject to an 

unacceptably high degree of variability reducing the value of the 

whole exercise. 

http://explorable.com/research-methodology
http://www.wisegeek.com/what-are-online-surveys.htm
http://explorable.com/selecting-the-survey-method


35 

Therefore, between these two extremes lies the most efficient (in cost 

terms) sample size for the given the study objective. Two major methods of 

sample size calculation are usually used, which are presented below. 

3.3.1 Standards of Bureau of Public Roads (BPR): 

The sample size can be determined based on the basis of population of the 

study area, and the standards given by the USA Bureau of Public Roads 

(BPR) are as shown in Table 3.1. 
 

Table 3.1: Standards of Bureau of Public Roads (BPR) for Sample 

Size Calculation 

Population of Study Area 
Sample size (Dwelling Units) 

Recommended Minimum 

Under 50,000 1 in 5 (20%) 1 in 10 (10%) 

50,000 – 150,000 1 in 8 (12%) 1 in 20 (5%) 

150,000 – 300,000 1 in 10 (10%) 1 in 35 (2.86%) 

300,000 – 500,000 1 in 15 (6.67%) 1 in 50 (2%) 

500,000 – 1,000,000 1 in 20 (5%) 1 in 70 (1.43%) 

Over 1,000,000 1 in 25 (4%) 1 in 100 (1%) 
Source: U.S Bureau of Public Roads, 1967 

3.3.2   Sample Size Statistical Formulas: 

The sample size can be calculated according to statistical formulas. These 

formulas can be found in most statistics textbooks, especially descriptive 

statistics dealing with probability. 

The sample size of infinite population can be calculated as: 

   
                 

  
 

Where: 

  : Sample size required. 

    
     : 

Standard normal statistic corresponding to (     confidence 

level.  



36 

 : Fraction of area under normal curve representing events not 

within confidence level (Thus,     is the desired level of 

confidence). 

P: Percentage of population picking a choice, expressed in decimal.  

C: Confidence interval or tolerable margin of error, expressed in 

decimal. 

The sample size of finite population is calculated as follows:  

           
 

         
 

Where: 

 :   Old sample size. 

P:   Population of the study area.  

Therefore, the sample size is calculated using the infinite population 

formula first. Then the sample size derived from that calculation is used to 

calculate a sample size for a finite population. 

3.4 Overview of Linear Regression Method 

There are several methods used in relating trips produced or attracted to the 

causal factors, which include the regression method and the cross-

classification method.  

Linear regression method assumes that observations of the magnitude of 

the dependent variable, Y (number of trips), can be obtained for n 

observations of explanatory variables (Xs), and that an equation of the 

form      is to be fitted to the data. This equation is for the single 



37 

independent variable case. The equation may be for independent variables 

as in the equation: 

                     

Estimating the best regression line depending on the coefficient of 

determination (R²), the t–test for the parameters, and the F-test can be done 

using computer programs like Excel, SPSS or XLSTAT. 

3.4.1 The Linear Regression Analysis Process  

Regression equations can be developed through the following sequence of 

steps (Hutchinson, 1974): 

1. Examining the relationships between the dependent variable and 

each of the independent variables in order to detect nonlinearities. 

If nonlinearities are detected, the relationship must be linearized by 

transforming the dependent variable, the independent variable, or 

both. 

2. Developing the inter-correlation matrix involving all the 

independent variables. 

3. Examining the simple correlation matrix in order to detect the 

potential sources of multicollinearity between pairs of the 

independent variables. 

4. After examining the correlation matrix, if there is multicollinearity 

between two independent variables (much closer to one), then one 

of them must be eliminated from the regression process. 



38 

5. After choosing the correlated independent variables, some of 

regression equations are suggested, and then the parameters of each 

of the potential regression equations are estimated. 

6. For every model built, relevant tests are conducted to assess the 

goodness of the model based on logic and statistics testing. The 

statistical tests include the coefficient of determination (R²), the t-

test for the parameters of the models, and the F-test for each model. 

In addition, there are logical aspects that must be taken in 

consideration, because the model must be valid statistically and 

logically.   

7. After all the models or equations are examined based on test 

results, the outcome is summarized. The best model is then chosen. 

3.4.2 Regression Model Building Approaches 

The most commonly used approaches in selecting the explanatory variables 

to be included in the regression model are the forward selection approach 

and the backward elimination approach. 

The forward selection approach starts with a model that contains only the 

constant term. At each step, the explanatory variable that results in the 

largest change in overall fitness is added to the model. This process is 

repeated until there is no more variable that results in a significant increase 

in overall fitness.  

The backward elimination approach starts with a regression model that 

contains all the explanatory variables. At each step, the explanatory 

variable that changes the overall fitness least is removed. The removal 



39 

process is repeated until removal of any variable results in a significant 

change in overall fitness. When deciding which explanatory variable 

should be removed from the regression model, the t-test is usually 

employed. Generally speaking, the explanatory variable with the least t-

value is selected for removal. 

In this research, the backward elimination approach is employed for 

building the trip generation models. 

3.5 Unit of Analysis 

Variables can generally be inferred from the unit of analysis. For travel 

modeling, two units are normally considered: the household and the 

individual. The household unit is preferred for various reasons: from a trip 

making point of view, the home is the basis where most trips start and end; 

from an economic point of view, income or car-ownership are usually 

shared by all members of the household; or from a social context, the 

family constitutes the „cell society‟ where all basic needs are usually met. 

Alternatively, if the individual is considered to be the base unit then the 

problem of allocation of some of the above mentioned variables needs to 

be overcome, or different quantities have to be taken into consideration. 

Therefore, the household is used as the unit of analysis in this research. 

3.6 Data Analysis Software 

There are many statistical softwares that can be used for data analysis. The 

Statistical Package for the Social Sciences (SPSS) is one of the most 

widely used statistical analysis packages. The SPSS program provides a 

wide range of procedures and tests used in statistics. Moreover, it offers 



40 

descriptive statistics such as frequencies, means, and correlations. Finally, 

it is useful for making charts. 

In this research, the SPSS software will be used to estimate the trip 

generation models using the linear regression analysis method. 

3.7 Model Specification 

The multiple regression analysis is one of the popular forms of model 

structure, which can be applied for trip generation models. Consequently, 

multiple regression equations will be used for developing the trip 

generation models for this study. The most common form of trip generation 

models is the linear function of the form 

                       

Where   is the constant term,    are the explanatory variables,   are the 

partial slope coefficients of the regression equation, and u is an error term 

that is assumed to be a random variable. The coefficients of the regression 

equation can be obtained by doing regression analysis using statistical 

analysis software. The above equation is called multiple linear regression 

equation. 

3.8 Models Estimation 

After reviewing the related literature, the trip generation models will be 

estimated. The models will be estimated using the multiple linear 

regression technique by regressing the dependent variable on each of the 

explanatory variables.  



41 

In this research, the trip generation models are divided into three 

categories. The first includes a general trip generation model. The second 

includes trip generation models according to trip purpose. The final 

category includes trip generation models according to trip time.  

3.8.1 General Trip Generation Model 

The general trip generation model will be estimated according to the 

following linear regression equation: 

                       

3.8.2 Trip Generation Models by Trip Purpose 

The work trip generation model, the education trip generation model, the 

shopping trip generation model, the social trip generation model, and the 

recreational trip generation model will be estimated according to the 

general form of linear regression equation: 

                        

Where i=1 to 5, covering the five trip purposes. 

3.8.3 Temporal Trip Generation Models 

Five temporal trip generation models will be estimated. These models 

include the trip generation model for trips made before 8 AM, the trip 

generation model for trips made between 8-9 AM, the trip generation 

model for trips made between 9 AM-12 PM, the trip generation model for 

trips made between 12-4 PM, and the trip generation model for trips made 

after 4 PM. The trip generation models according to timing will be 

estimated according to the following equation: 



42 

                        

Where k=6 to 10, covering the five periods under consideration. 

3.9 Statistical Tests 

The most common statistical tests that are used in model selection process 

are discussed below. 

3.9.1 Correlation Matrix and VIF: Testing for Multicollinearity 

The problem of multicollinearity arises when two or more explanatory 

variables included in the regression model have linear relationships. There 

are two types of multicollinearity: exact or perfect multicollinearity and 

inexact or imperfect multicollinearity.  

If two or more independent variables have exact or perfect linear 

relationships, then we have exact or perfect multicollinearity. On the other 

hand, inexact multicollinearity occurs in the case that two or more 

independent variables are highly correlated. 

In the case of exact multicollinearity, there is no unique solution to the 

normal equations derived from the least squares principle. As a result, the 

regression coefficients cannot be estimated. By contrast, when explanatory 

variables have less than exact linear relationships, the normal equations can 

usually be solved to yield unique estimates.  

However, there are some consequences. First, regression coefficients are 

still best linear unbiased estimators (BLUE). Second, it is more difficult to 

precisely identify the separate effects of the correlated explanatory 



43 

variables. Finally, variances and standard errors are usually higher, making 

t-statistics lower and possibly insignificant.  

Analysis often relies on what is called the Variance Inflation Factor (VIF) 

to detect multicollinearity more formally. The VIF shows how the variance 

of an estimator is inflated by the presence of multicollinearity.  

As R
2
, the coefficient of determination of a given explanatory variable with 

other remaining explanatory variables in the model, increases toward one, 

the VIF also increases. The larger value of VIF, the greater the degree of 

muticollinearity of one explanatory variable with the other explanatory 

variables. As a rule of thumb, if the VIF of a variable exceeds 10, that 

variable is said be highly collinear (Kleinbaum et al., 1988). 

3.9.2 R-Squared: Goodness of Fit 

The R-squared (R
2
), also known as the coefficient of determination, 

measures the goodness of fit of the regression model. It measures the 

proportion of the total variation in the dependent variable that can be 

explained by the explanatory variables included in the model. The value of 

R-squared lies between 0 and 1. A value of R-squared closer to 1 indicates 

that the model has good fit, whereas a value closer to 0 indicates that the 

model has poor fit. 

However, there is no standard on how high R
2
 value is “good” enough. It 

depends on the application.  

The R-squared is calculated as follows: 

   
                               

                          
 



44 

The ANOVA test results are used to show the analysis of the total variance 

in the dependent variable. This variance is divided into two sources: 

variance due to regression and variance due to errors. 

3.9.3 F-Test: Testing Overall Significance of Model 

The F-statistic is used to test whether the regression coefficients are jointly 

equal to zero or not. In other words, the F-test is used to test the overall 

significance of the regression model.  

The null hypothesis for testing the overall significance of the model is that 

the regression coefficients for the explanatory variables are all equal to 

zero. The alternative hypothesis is that at least one of these coefficients is 

not equal to zero. Usually, a 95% level of significance for the F-value is 

accepted. 

Table 3.2 is a format of ANOVA test results table. 

 

Table 3.2: ANOVA Test Results Table 

Source 
Sum of 

Squares 

Degrees of 

Freedom 

Mean 

Square 
F-Statistic Significance 

Regression      

Residual      

Total      

3.9.4 T-Test: Testing Individual Coefficients 

The t-statistic is used to test the significance of individual regression 

coefficients. As a rule of thumb, if the calculated t-statistic is greater than 

two in absolute value, it is concluded that the estimate is statistically 

different from zero at the 95% level of significance. 

 

 



45 

3.10 Logical Aspects Used in Model Selection 

The following is a list of logical aspects that should be taken into 

consideration in selecting the best regression model: 

1. The regression coefficients must have the correct expected sign and 

their magnitudes must be reasonable. For example, an explanatory 

variable is expected to have a positive effect on the dependent 

variable, whereas the result of regression analysis shows a negative 

effect. 

2. The constant (intercept) term must be reasonable both in value and 

sign. 

3.11  Determination of Study Area 

Urban transportation planning means planning the transportation facilities 

based on the potential of growth of urban area for the next coming decades. 

So, the study area must cover the existing and potential continuously built-

up areas of the city.  

The study area is specified by imaginary line representing the city 

boundaries, called external or outer cordon line. There are four factors for 

selecting and fixing this cordon line (Arasan, 2012): 

1. The external cordon line should cover the built-up areas and the 

areas that will be developed in the future during the planning 

period. Figure 3.1shows the Draft Master Plan of Jericho City for 

the year 2013.There is an imaginary line representing the approved 

expansion areas, and depending on this line, the external cordon 

line is fixed as the boundary of the study area. 

2. The external cordon line should include all areas of systematic 

daily life of people, oriented towards the city center and should in 



46 

effect be the commuter shed. Because there are areas beside 

Jericho City that include systematic daily life and affect travel 

patterns, these areas are included in the external cordon line. These 

areas are Al-Nwe'ma and Ein Dyok Alfoqa. 

3. The external cordon line should be continuous and uniform in its 

course, so the movements cross it only in one point. The line 

should intersect roads where it is safe and convenient for carrying 

out traffic surveys. 

4. The external cordon line should be compatible with the previous 

studies of the area or studies planned for the future. These previous 

studies could be related to census operation or population census. 
 

 

Figure 3.1: Draft Master Plan of Jericho City (2013) 

Source: Jericho Municipality, 2013. 



47 

3.12 Zoning System  

Zoning means dividing the study area after defining the boundary into 

smaller land use areas called traffic analysis zones (TAZ) or simple zones. 

The zones within the study area or external cordon line are called internal 

zones. The division of internal zones and their numbers will depend on a 

compromise between a series of criteria discussed in this section. The 

region external to the study area is normally divided into a number of 

external zones. In some cases, as in this case, it might be enough to 

consider each external zone to represent the rest of the surrounding regions 

in a particular direction; the boundaries of these different slices of the rest 

of the surrounding regions could represent the natural catchment areas of 

transport links feeding into the study area (Ortuzar and Willumsen, 1996). 

Zones are modeled as if all their attributes and properties were 

concentrated in a single point called the zone centroid. The centroids are 

connected to the nearest road junction or transport station by centroid 

connectors. The centroid and the centroid connectors are notional and it is 

assumed that all people have same travel cost from the centroid to the 

nearest transport facility which is the average for a zone (Ortuzar and 

Willumsen, 1996). 

The following are the general guidelines for zoning process (Ortuzar and 

Willumsen, 1996): 

1. The zoning system must be compatible with other administrative 

divisions, particularly with census zones, this is probably the 



48 

fundamental criterion and the rest should only be followed if they do 

not lead to inconsistencies with it. 

2. Zones should be as homogenous as possible in their land use and 

population composition, so as to reflect accurately the associated trip 

making behavior. There are different types of land uses in urban 

areas including residential, industrial, institutional, and recreational. 

3. The anticipated changes in land use should be considered when sub-

dividing the study area into zones. 

4. Zone boundaries must be compatible with cordons and screen lines 

and with those of previous zoning systems. In this study, major roads 

are considered as zone border and commuter shed for particular 

traffic zone. On the other hand, external cordon lines are considered 

as part of boundaries of traffic zones, and wadis or rivers are also 

considered as zone boundaries. 

5. The shape of the zones should allow an easy determination of their 

centroid connectors. This is important for later estimation of intra-

zonal characteristics, and important for determination of the 

centroid, which represents the origin and destination of travel. 

6. Zones do not have to be of equal size, if anything, they could be of 

similar dimensions in travel time units, therefore generating smaller 

zones in congested than in uncongested areas. 

7. In general, zones should not be too large or too small, because the 

too large zone causes considerable errors in data and the too small 

zone causes difficulty in handling and analyzing the data. As a 

general guide, a population of 1,000-3,000 may be the optimum of 



49 

small area, a population of 5,000-10,000 may be the optimum of 

large urban areas, and as a whole the size of each TAZ between 0.25 

to 1 squared mile, or 0.5 to 2.5 square kilometer. Also, no generated 

more than 15,000 person trips in base and future years (Cambridge 

Systematics, Inc., 2007). 

8. Natural or physical barriers such as canals, rivers, and so on can 

form convenient zone boundaries. In this case, there are two rivers 

which can be considered as screen lines and zone boundaries as 

shown in Figure 3.2. (Cambridge Systematics, Inc., 2007). 

 

Figure 3.2: Major Roads and Natural Barriers in Jericho City Used for Specifying 

Zones Boundaries 



50 

9. Zone boundaries should preferably be watersheds of trip making 

(Arasan, 2012). If there are arterial roads, and there are a lot of 

commuters towards that arterial road for commuting that joining 

area, then the adjoining area becomes a commuter shed for this 

particular road, like the catchment area and the watershed (Arasan, 

2012). Explanation of commuter shed is shown in Figure 3.3. 

Arterial roads in Jericho City, which are considered as a commuter 

shed, are shown in Figure3.3. 

 
Figure 3.3: Explanation of Commuter Shed 

The purpose of zoning system is to: 

1. Facilitate spatial quantification of land use and economic factors, 

which influence travel pattern (Arasan, 2012). This can be done by 

matching the land-use plan with the boundaries of traffic zones as 

shown in Figure 3.4. 

After matching the land-use plan with zone boundaries, the area and land-

use of each traffic zones are specified for this research as shown in Table 

3.3. 

 



51 

 

Figure 3.4: Land Uses Plan with the Boundaries of Traffic Zones 

Source: Master Plan of Jericho (Jericho Municipality, 2010) 

 

 

 

 

 

 

 

 

 

 

 
 
 



52 

Table 3.3: Land Uses and Areas for Traffic Zones 

Traffic Zone 

No. 

Area 

(Km²) 
Land Uses in Traffic Zone 

1 0.547 Residential, commercial, public facility. 

2 2.479 
Residential, commercial, public facility, 

agricultural, heritage. 

3 2.409 Residential, agricultural, heritage. 

4 1.254 
Residential, commercial, public facility, 

parking, industrial, agricultural. 

5 3.726 
Residential, agricultural, commercial, heritage, 

cemetery. 

6 2.049 
Residential, agricultural, commercial, public 

facility, park. 

7 1.668 
Residential, agricultural, commercial, public 

facility, park, heritage. 

8 2.142 
Residential, agricultural, commercial, public 

facility. 

9 2.113 Residential, agricultural, public facility. 

10 4.185 
Residential, commercial, public facility, 

agricultural. 

11 1.262 
Residential, commercial, public facility, 

agricultural, cemetery. 

12 3.864 
Residential, agricultural, public facility, 

cemetery, parking, industrial, commercial. 

13 10.182 
Residential, agricultural, public facility, 

commercial. 

14 3.005 
Residential, public facility, commercial, 

parking, agricultural. 

Source: Master Plan of Jericho (Jericho Municipality, 2010) 

2. Help in geographically associating the origins and destinations of 

travel (Arasan, 2012).  

According to all previous points, the final TAZ's adopted for Jericho City 

are shown in Figure 3.5. 



53 

 
 

Figure 3.5: TAZ's of Jericho City 

 

 

 

 

 

 

 

 

 



54 

 

 

 

 

 

 

 

Chapter Four 

Field Survey and Data Collection 

 

 

 

 

 

 

 

 

Chapter Four 



55 

Field Survey and Data Collection 

4.1 Population of Study Area 

In general, to specify the population of study area, statistics from the 

Palestinian Central Bureau of Statistics (PCBS) will be used. According to 

the specified borders of study area, there are refugee camps and villages 

outside the borders of Jericho Municipality, but these camps and villages, 

which are located just beyond the municipality borders have systematic 

daily life and affect the travel pattern of Jericho City. On the other hand, 

there are villages considered as part of Jericho and Valleys Governorate, 

but these villages will not be considered in this research because they are 

far from Jericho City.  

In order to calculate the sample size that will be used in this study, first the 

population size is specified from the PCBS as appears in Table 4.1. 

The population that will be considered in this study is 35,885 people. 

Based on a random sample of 100 household respondents, the average 

household size in the study area is calculated to be 5.03 persons. The 

number of households in the study area is thus 7,134households 

(35,885/5.03). 

 

 

 

Table 4.1: Localities in the Study and Estimates of Population, 

2013 



56 

Population 
Locality 

Type 

Locality 

Code 
Locality Name 

21,411 Urban 351920 Jericho City 

1,453 Rural 351840 Al New'ma 

958 Rural 351845 Ein Dyok Alfoqa 

3,688 Refugee Camp 351865 Ein Al Sultan Refugee Camp 

8,375 Refugee Camp 351975 Aqbat Jaber Refugee Camp 

21,411 
  

Total urban  

2,411 
  

Total rural 

12,063 
  

Total refugee camps 

35,885 
  

Total of Study Area 

Source: The Palestinian Central Bureau of Statistics, 2013. 

4.2 Sample Size of Study Area 

According to the standards of the USA Bureau of Public Roads (BPR, 

1967), the minimum sample size that will be taken in this study is 1 in 10 

of dwelling units since the population of the study area is under 50,000.The 

sample size that will be considered in this study is 10% of the total number 

of households in the study area, 713 households (10% times 7,134), 

distributed over 14 TAZ‟s according to percentages and numbers shown in 

Table 4.2. 

In order to calculate the sample size based on Statistical Formulas, a 

random sample of 100 household respondents is initially taken to calculate 

the percentages of population (P) that make different types of trips. Then, 

these percentages are used in the equation to calculate the actual sample 

size required for this study as shown in Table 4.3. 

 

Table 4.2 Number of Households per Traffic Zone for Study Area 

and Sample Size Required 



57 

Traffic Zone 

No. 

# of Housing 

Units *  
Population ** 

% of Housing 

Units 

Sample 

Size 

1 188 946 3% 19 

2 489 2458 7% 49 

3 190 958 3% 19 

4 733 3,688 10% 73 

5 289 1,453 4% 29 

6 647 3256 9% 65 

7 560 2816 8% 56 

8 606 3047 8% 61 

9 55 275 1% 5 

10 301 1512 4% 30 

11 580 2915 8% 58 

12 162 814 2% 16 

13 632 3179 9% 63 

14 1703 8,838 24% 171 

Total 7,134 35,886 100% 713 
*Number of Housing Units in every traffic zone was found depending on the field 

survey. 

** Population in every traffic zone was found by multiplying the number of housing 

units by the average size of household (5.03 person/household). 

 

Table 4.3: Sample Size Calculation According to Statistical 

Formulas 

Trip Type P 1-P α Z c Ss 
Modified 

S 

Work 0.22 0.78 0.05 1.96 0.04 412 407 

Education 0.29 0.71 0.05 1.96 0.04 494 488 

Social 0.17 0.83 0.05 1.96 0.04 339 336 

Shopping 0.24 0.76 0.05 1.96 0.04 438 433 

Recreational 0.09 0.91 0.05 1.96 0.04 197 196 

Total 1.00 
   

 1,880 1,859 

So, the sample size according to the statistical formula is 370households 

(1,859/5.03). However, as the sample size calculated from the BPR is 

found to be 713, which is higher than the 370 households, the later will be 

adopted in this study. This higher sample size is adopted to get a higher 

degree of accuracy. 



58 

4.3 Sampling Method 

The sample that is drawn from the household population is a stratified 

random sample that includes 713 households. Stratification reflects 

geographic locations defined as TAZ. Therefore, the study area is divided 

into 14 TAZ's. The samples are randomly selected from each of the 14 

TAZ's. The sample size of each TAZ is shown in Table 4.2. 

4.4 Questionnaire Design 

In terms of layout, the order of the questions normally seeks to minimize 

the respondent‟s resistance to answering them; the survey instrument (and 

any personal interviews) should try to satisfy the following criteria 

(Ortuzar and Willumsen, 1996): 

1. The questions should be simple and direct. 

2. The number of open questions should be minimized, so long 

answer avoided and use of code for answer preferable. 

3. Difficult questions (e.g. relating to income) are usually put at the 

end. 

4. Travel information must include the purpose of the trip. 

5. Because of the growth of non-car modes, seek information about 

all modes of travel, including non-motorized travel. 

6. Due to the growing importance of independent trips by children, 

and of non-motorized modes, all people in the household should be 

included in the survey, including non-family members, like maids 

in developing countries. 



59 

7. To facilitate the respondents task of recording all travel, it is 

recommended that an activity-recall framework be used; people 

record travel in the context of activities they have undertaken 

rather than simply trips they have made; this has been shown to 

result in much more accurate travel measurement. 

8. Since people have difficulty recalling infrequent and discretionary 

activities, even when they are recent, it is recommended that travel 

day or days be assigned to each household in advance, they should 

be given a brief diary in advance of these days: The information in 

the diary may be then transferred to the self-completion form or 

reported to the interviewer at the end of the day. Anyhow, in this 

study, the recalling of trips or activities depended on the memory 

of persons in the household. 

4.5 Required Information 

For any type of household survey, it is recommended that the survey be 

divided into two parts: 

1. Personal and household characteristics and identification: This part 

includes questions designed to classify the household members 

according to their relation to the head of household (e.g. wife or 

son), sex, age, possession of a driving license, educational level, and 

occupation. In order to reduce the possibility of a subjective 

classification, it is important to define a complete set of occupations 

(non-household surveys are usually concerned only with person 

being interviewed; however, the relevant questions are the same or 



60 

very similar). This part also includes questions designed to obtain 

socio-economic data about the household, such as characteristics of 

the house, identification of the household vehicles, house ownership 

and family (Ortuzar and Willumsen, 1996). 

2. Trip data: This part of survey aims at detecting and characterizing all 

trips made by the household members identified in the first part. A 

trip is defined as any movement outside a building or premise with a 

given purpose; but the information sought considers trips by stages, 

where a stage is defined by a change of mode (including walking). 

Each stage is characterized on the basis of variables such as origin 

and destination (normally expressed by their nearest road junction), 

purpose, start and ending times, mode used, amount of money paid 

for the trip, and so on (Ortuzar and Willumsen,1996). 

According to the points mentioned in the questionnaire design and the 

required information, the questionnaire form used in this thesis is shown in 

Appendix B. This questionnaire aims to collect in the first part personal 

and identification characteristics of households, which are considered as 

explanatory variables. These variables are used in the attempt to build the 

models, which will be examined statistically and logically to adapt the 

most appropriate model that can predict the trips generated by the TAZ's.  

The questionnaire contains two tables. The first table gives details about 

the socio-economic variables of households. Such details contain: gender, 

age, possession of driving license, possession of transportation facilities 

like private cars, bicycles, and motorcycles, house type (independent home 



61 

or apartment), and household income. The bicycles are taken into 

consideration because Jericho City is flat and there are bikes in almost 

every household. 

The explanatory variables that have been considered for possible use in the 

modeling process are summarized in Table 4.4. 

The age variable is divided into five groups as shown in Table 4.4. The 

first age group of ages up to 16 years includes persons who are assumed to 

be in schools. The second age group of ages between 17  to 30 includes 

persons who are either continuing education in universities or part of the 

labor force. The third and fourth age groups include persons who are heads 

of families and part of the labor force, which are divided into an age group 

from 31 to 50 and another from 51 to 64. Finally, the fifth age group of 

ages 65 and above includes elderly persons who are usually retired.    

The second table is designed to collect data concerned with the second part 

of survey (trip data). This table is made to calculate the total trips generated 

by the household and the number of trips for each type of trips (work trips, 

education trips, shopping trips, social trips, and recreational trips), and the 

trips according to their time. These trips are divided into five groups. The 

first group includes trips made before 8 AM. Trips made during this period 

are called the morning peak period trips which are usually made by most 

students and employees. The second time group includes trips that are 

made between 8-9 AM. During this time period, some students and 

employees make their trips. The third and fourth time periods include trips 

made between 9 AM -12 PM and 12-4 PM, respectively. These time 

groups are usually considered to be the AM and PM off peak periods. The 



62 

fifth time group includes trips made after 4 PM. This time period is 

considered to be the evening peak period. Usually, most of shopping, 

social, and recreational trips are made during this time period. All these 

trips are summarized as shown in Table 4.5. 
  

Table 4.4 : Explanatory Variables Used in the Models 

Number of persons in the household X1 

Number of males in the household X2 

Number of females in the household X3 

Number of employed persons in the household X4 

Number of persons receiving education in the household X5 

Number of persons under 16 years in the household X6 

Number of persons between 17 and 30 years in the household X7 

Number of persons between 31 and 50 years in the household X8 

Number of persons between 51 and 64 years in the household X9 

Number of persons above 65 years in the household X10 

Number of licensed drivers in the household X11 

Number of cars owned by a household X12 

Number of bicycles owned by a household X13 

Number of motorcycles owned by a household X14 

Monthly household income (Thousand New Israeli Shekel) X15 

House type: 1if  Independent, 0 if Apartment X16 

 

Table 4.5: Dependent Variables Used in the Models 

Number of daily trips made by household Y 

Number of daily work trips made by household Y1 

Number of daily educational trips made by household Y2 

Number of daily shopping trips made by household Y3 

Number of daily social trips made by household  Y4 

Number of daily recreational trips made by household  Y5 

Number of daily trips made by household before 8 AM Y6 

Number of daily trips made by household between 8-9 AM Y7 

Number of daily trips made by household between 9-12 AM  Y8 

Number of daily trips made by household between 12 AM - 4 PM Y9 

Number of daily trips made by household after 4 PM Y10 

4.6 Conducting Field Survey  



63 

After determining the sample size, designing the questionnaire, and 

choosing the method of survey, the field survey in this study was 

conducted according to the following steps (Mathew, 2011): 

1. Two skilled enumerators are chosen to help in the survey process.  

2. Enumerators have to be trained by briefing them about the details of 

the survey and how to conduct the survey.  

3. The enumerators will be given random household addresses. 

4. The random sample of households in different traffic zones will be 

numbered on a printed map and the same numbers are put on the 

questionnaire layout, where each number on the map and the 

questionnaire represents one household. 

5. Since the actual survey may take place any time during the day, the 

respondents are required to answer the question about the travel 

details according to the previous day. This previous day must be 

typical working day because working days represent the most critical 

situation. In Palestine, the typical working days are from Sunday to 

Thursday.  

6. The enumerators have to first get permission to be surveyed from the 

household. Then, they ask the household members about the details 

required in the questionnaire. Each member of the household should 

give answers about his or her own travel details, except for children 

below 12 years. Trip details of children below 5 years are normally 

ignored. 

 



64 

 

 

 

 

 
 

 

Chapter Five 

Data Analysis and Results 

 

 

 

 

 

 

 

 

 

 

Chapter Five 



65 

Data Analysis and Results 

5.1 Descriptive Data 

In this section, descriptive statistics for both the dependent variables as 

well as the explanatory variables will be shown and discussed. This section 

is intended to give an overview of the distribution of various trips 

according to their purpose as well as to their time. This shows which trips 

are the most frequent.  

Moreover, this section includes figures that relate the number of daily trips 

with each of the explanatory variables so as to show graphically if these 

variables are related or not. This also shows if these relationships are linear 

or not.  

5.1.1 Descriptive Data of Dependent Variables 

Table 5.1 lists discriptive statistics for the total daily household trips. There 

has been a total of 4,913 trips made by 713 households. 
 

Table 5.1: Descriptive Data for the Total Daily Household Trips 

Total Trips Mean 
Standard 

Deviation 
Maximum Minimum Range 

Y 6.89 3.096 16 0 16 

It is apparent from Table 5.1 that the average daily trips per household is 

around 7. The maximum number of daily trips per household is 16 whereas 

the minimum number is 0. 

The discriptive statistics for the daily household trips according to their 

purpose are shown in Table 5.2. 

 

Table 5.2: Descriptive Data for the Daily Household Trips by 

Purpose 



66 

Trip Purpose Mean 
Standard 

Deviation 
Maximum Minimum Range 

Work 1.60 0.817 5 0 5 

Education 1.84 1.575 6 0 6 

Shopping 1.52 0.996 7 0 7 

Social 1.03 1.143 7 0 7 

Recreational 0.92 1.450 6 0 6 

The discriptive statistcs for the temporal daily household trips according 

are shown in Table 5.3. 

Table 5.3: Descriptive Data for the Daily Household Trips by Time 

Trip Time Mean 
Standard 

Deviation 
Maximum Minimum Range 

Before 8 AM 2.58 1.640 7 0 7 

8-9 AM 0.57 0.820 5 0 5 

9-12 AM 0.24 0.501 3 0 3 

12 AM - 4 PM 0.54 0.653 4 0 4 

After 4 PM 2.95 1.820 10 0 10 

Classification of trips according to their purpose is important to be carried 

out since people make the trips for various reasons. In this research, among 

4,913 trips generated by 713 surveyed households, most of the trips (27%) 

are education trips. Work trips account for 23% of total trips generated. 

The distripution of household trips according to purpose is shown in Table 

5.4. The same data are shown graphically in Figure 5.1. 

Table 5.4: Distribution of Daily Household Trips by Purpose 

Trip Purpose Number % 

Work 1,141 23% 

Education 1,310 27% 

Shopping 1,085 22% 

Social 727 15% 

Recreational 650 13% 

Total 4,913 100% 
 



67 

 

Figure 5.1: Distribution of Daily Household Trips by Purpose 

People do their activities in different times within a day. Hence, it is 

essential to carry out the temporal distribution of trips generated. 

According to the survey data, most of the trips are distributed within peak 

hours where 37% of trips are made during morning peak hours (before 8:00 

AM) and 43% are made during afternoon peaks (after 4:00 PM). The 

temporal distribution of daily household tripsis shownin Table 5.5. The 

same data are presented graphically in Figure 5.2. 

Table 5.5: Temporal Distribution of Daily Household Trips 

Trip Time Number % 

Before 8 AM 1,841 37% 

8 AM - 9 AM 409 8% 

9 AM - 12 PM 174 4% 

12 PM - 4 PM 383 8% 

After 4 PM 2,106 43% 

Total 4,913 100% 

23% 

27% 

22% 

15% 
13% 

0%

5%

10%

15%

20%

25%

30%

Work Trips Education
Trips

Shopping
Trips

Social Trips Recreational
Trips



68 

 
 

Figure 5.2: Temporal Distribution of Daily Household Trips 

 

5.1.2 Descriptive Data of Explanatory Variables 

The descriptive statistics for the household size are shown in Table 5.6. 

 

Table 5.6: Descriptive Data for Household Size 

Mean 
Standard 

Deviation 
Maximum Minimum Range