The Appication Of Alternative Method

Published Date: 02 Nov 2017

CHAPTER 4

This chapter presents the application of the alternative method for estimating the value of time, value of reliability and route choice decision of freight operators in logistic firms in Thailand. The following research was designed, and had data collected and analyzed from the research framework in chapter 3.

4.1 Research Design

This part explains the process of stated preference design in the initial research, the Revealed Preference design (pilot survey), the experimental design, the questionnaire design, the web survey design and main survey (SP survey) that’s shown in the framework below.

Survey Plan

Initial research

Revealed preference design

Redesign

First pilot survey

Experimental and Questioners design

Redesign

Second pilot survey

Web survey design

Main survey

Figure 4.1 Research design Framework

4.1.1 Initial Research

The principal survey tool was web base Adaptive Stated Preference (ASP) data collection. The survey plan was separated into three surveys, which are the first pilot survey, the second pilot survey and the main survey respectively. The purpose of the first pilot survey was to estimate the approximate value of each attribute from the revealed preference (SP) questionnaire. The second pilot survey was purposed to find the best fit for choice experiments from full choice experiment survey. Finally, the purpose of the main survey was to estimate the discrete choice modeling and attribute valuation (VOT and VOR) using the web base Adaptive Stated Preference survey.

4.1.2 Revealed Preference Design

The revealed preference (RP) deals with actual data, which is the best way to get real data at present. Revealed preference (RP) uses direct questions, such as: gender, age, occupation, travel time, toll cost, time reliability etc.

The RP survey is a very useful method for approximating the real value of each attribute that’s used--travel time, toll cost and travel time reliability in trips from the Bangkok port (Bangkok) to the Map Ta Phut sea port (Rayong). The survey was separated into two sections.

Respondent characteristics and company background; gender, age, occupation, education, company location, age work.

Route choice from Bangkok port (Bangkok) to Map Ta Phut sea port (Rayong); e.g. travel time, toll cost, travel time reliability.

Details of variables, operational definition and level of measurement were shown in Table 4.1.

Table 4.1 Variable, operational and level of measurement

Variable

Operational

Level of measurement

Name

Normal

Gender

Male

Female

Normal

Age

Age (year)

Ratio

Response education

Less than bachelor degree

Bachelor degree

Master degree

Doctor degree

Normal

Response company

Normal

Response position

Normal

Response company location

Normal

E-mail

Normal

Phone no.

Normal

Work experience

Normal

Travel time

Ratio

Travel cost

Ratio

Travel time reliability

Ratio

In the first section the purpose was to get general information of the respondents and their business: name, age, gender education. Furthermore, their contact information: email, addresses and phone number was recorded into the database. The contact information is very useful when starting to collect information for the main survey; the web base survey will send information directly to their personal email, so this can increase the response rate. In the second section of the RP survey, the purpose is to get the respondents experience regarding freight trips from Bangkok port (Bangkok) to Map Ta Phut sea port (Rayong) using three route; (1)Motorway highway no.7, (2)Bhuraphavithe expressway and (3)Bangna Trad road,. in terms of travel time, toll cost and travel time reliability on each route.

Their revealed preference is shown in the appendix A. The survey was run from the 15th to the 31st of December, 2012 for about 15 days of data collection by using the web base survey (webserv.kmitl.ac.th/surveysp/). Figure 4.2-4.3 shows the web base short screen. The website was sent to 67 logistic companies by email. Seven companies out of 67 companies contacted responded, for a response rate of approximately 10 percent.

Figure 4.2 RP web survey screen short section 1

Figure 4.3 RP web survey screen short section 2

Table 4.2-4.3 illustrates the revealed preference data (pilot survey) of logistic managers results: (1) respondent characteristics and company background and (2) Route choice from the Bangkok port (Bangkok) to the Map Ta Phut sea port (Rayong).

Table 4.2 Result of Logistic Manager and Firm Characteristics.

Variable

Frequency

Percentage (%)

Gender

Male

Female

7

5

58.33

41.67

Age

Less than 20 years old

20-30 years old

31-40 years old

41-50 years old

More than 50 years old

0

1

5

1

5

0.00

8.33

41.67

8.33

41.67

Education

Less than bachelor degree

Bachelor degree

Master degree

Doctor degree

0

8

4

0

0.00

66.67

33.33

0.00

Job position

Manager

Logistic manager

3

9

25.00

75.00

Company location

Bangkok

Other

9

3

75.00

25.00

Table 4.3 Result of Route Choice from Bangkok Port (Bangkok) to Map Ta Phut Seaport (Rayong)

Variable

Minimum

Maximum

Average

Standard diversion (SD)

(1) Motorway highway no.7

Travel time, (hour)

Toll cost, (baht)

Travel time reliability (%)

2:30

300

95

2:45

400

90

2:35

357.14

92.14

0:07

44.99

2.67

(2) Bhuraphavithe expressway

Travel time, (hour)

Toll cost, (baht)

Travel time reliability (%)

2:30

300

85

3:30

350

90

2:51

314.28

87.85

0:24

24.40

2.67

(3) Bangna Trad road

Travel time, (hour)

Toll cost, (baht)

Travel time reliability (%)

3:30

0

75

4:00

0

80

3:46

0

77.85

0:13

0

2.67

Total

Travel time, (hour)

Toll cost, (baht)

Travel time reliability (%)

2:30

0

75

4:00

400

95

3:02

223.81

85.95

0:32

165.54

6.64

Logistic Managers and Firm Characteristics

The results in Table 4.2 show that most respondents were male (58.33%), their average age was more than 30 years old (91.67%), most graduated with at least a bachelor’s degree (66.67%), most of respondents were logistic managers (75.00%) and most of companies are located in Bangkok area (75.00%).

Route Choice from Bangkok Port to Map Ta Phut Seaport

The results in Table 4.3 show that the average travel times were 2:35, 2:51, 3:46 hours on route Motorway No. 7, the Bhuraphavithe expressway, and the Bangna Trad road respectively. The average travel time on all three routes was 3:02. And the average toll costs for all three roadways was about 224 baht. In addition, Motorway highway no.7 had a higher value in travel time reliability and certainty than the other routes. Thus, respondents were usually willing to pay more to take Motorway highway no.7 in order to save travel time, which is shown by their willingness to pay 357.14 Baht on average when using this route.

4.1.3 Experimental Design

Willingness to pay attributes is considered for estimating the value of time and reliability of freight operators. To present the operational characteristics of freight trips three attributes (travel time, toll cost and travel reliability) were identified in the stated preference experiments. The optimal level of each attribute varied around the values for current trip for any given route scenario, such as: carrying goods or products from the Bangkok port to the Map Ta Phut seaport, which were determined using the RP survey. Thus, the levels for each attribute is shown in Table 4.4.

Table 4.4 Level of attribute

Level

Travel time (hr.)

Toll cost (Baht)

Travel time reliability (%)

L1

L2

L3

L4

L5

2:30

3:00

3:30

4:00

0

250

300

350

400

80

85

90

95

The chosen attributes and levels produced a full factorial design with 4251 = 80 possible choice experiments. The full factorial design is shown in the appendix B. The SPSS program was used to reduce number of choice experiments. Choice experiments reduced the number of route alternatives from 80 to 26 appropriate route choices which are shown in Table 4.5.

Table 4.5 Route Choice Experiment

Choice

Travel time (hr.)

Toll cost (Baht)

Travel time reliability (%)

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

3:00

2:30

2:30

4:00

3:30

4:00

3:00

3:30

4:00

2:30

3:00

3:30

2:30

3:30

2:30

2:30

2:30

3:00

2:30

3:30

4:00

2:30

3:30

4:00

2:30

2:30

300

350

0

400

250

250

250

300

350

250

0

400

400

0

400

350

300

400

0

350

0

0

0

0

250

300

90

85

90

90

95

85

80

80

80

90

95

80

80

85

95

95

85

85

80

90

80

95

95

90

95

95

4.1.4 Questionnaire Design

The questionnaire design for the main survey (ASP survey) was separated into three main parts.

Respondent characteristics and company background; e.g. gender, age, occupation, education, company location, age work.

Choice game; choice experiment.

Route choice decision; e.g. departure time, truck load, travel distance.

Details of variables, operational definition and level of measurement are shown in Table 4.6.

Table 4.6 Variable, operational and level of measurement

Variable

Operational

Level of measurement

Gender

Male

Female

Normal

Age

Age (year)

Ratio

Response education

Less than bachelor degree

Bachelor degree

Master degree

Doctor degree

Normal

Response company location

Normal

Type of goods carried

Chemicals

Fibers

Machines

Furniture

Wood

Textiles/clothing

Building/cement

Food

Metal/steel products

Automobile

Electrical/electronic

Paper/paper products

All goods

Normal

Business Classification

Truck shippers

Third Party Logistics

Cartage Company

Specialized Company

Normal

Travel time

Time

Ratio

Toll cost

Baht

Ratio

Travel time reliability

Percentage

Ratio

Total kilometers drive in a year

Less than 50,000 km.

50,001-100,000 km.

100,001-150,000 km.

150,001-200,000 km.

200,001-250,000 km.

250,000-300,000 km.

More than 300,000 km.

Normal

Types of facilities of the most frequently transport goods

Warehouse/Distribution Facility

Rail facility

Airport

Port

Customer facility

Normal

Average driver’s work years

Less than 1 year

1-2 years

3-5 years

6-10 years

11-20 years

21-30 years

More than 30 years

Normal

Number of truck drivers

Less than 10 persons

10-50 persons

51-100 persons

101-150 persons

151-200 persons

More than 200 persons

Normal

Number of trucks in fleet

1-5 trucks

6-20 trucks

21-50 trucks

51-100 trucks

101-200 trucks

More than 200 trucks

Normal

Decision maker

Driver has full control

Logistic manager has full control

Driver and logistic manager have the same amount of control

Normal

Typical environment operate

Almost entirely rural

Mostly rural with some urban

Equal amounts rural and urban

Mostly urban with some rural

Almost entirely urban

Normal

How is route determined?

Familiar route

Navigation system (GPS)

Consult a paper map

Talk with driver

Normal

Toll payer

Truck driver

Logistic company

Normal

Penalty for truck driver

Yes

No

Normal

Compensation of consumer

Yes

No

Normal

Normally weigh carrier

Empty truck

Less than truck load

Truck load

Normal

Normally freight distance

Shorter route (<150 km.)

Medium route (150-250 km.)

Longer route (>250 km.)

Normal

Normally freight time

Morning time

Afternoon time

Evening time

Night time

Normal

Logistic Manager and Firm Characteristic

The purpose of the first section was to get general information about respondents and their business. Questions were asked about their: age, gender, education, job position, company’s location, type of goods carrier and business classification.

Choice Game

This section’s purpose was to estimate the value of time and reliability by using the stated preference with choice experiment. The scenario is; "If your company has to carry goods or product from Bangkok port to Map Ta Phut sea port, which route you will take or avoid." The choice game design was shown in session 4.1.3 "Experiment design".

Toll cost is negatively affected to the utility of using toll road. If toll costs increased, the utility of using toll road will be reduced.

Travel time is negatively affect to the utility of using toll road. If travel time increased, the utility of using toll road will be reduced

Travel time reliability is positively related to the utility of using toll roads. If travel time reliability increased, the utility of using toll road will be increased.

Route Choice Decision

Goods movement businesses (all logistics companies) are some of the most ardent critics of using tolls to address the nation’s congestion, environmental, and roadway transportation infrastructure needs. Thus, a section was needed to understand how goods movement businesses assess tradeoffs in using or avoiding tolled facilities. This research considered the following factors that might impact their decisions. The expected impacts of those variables with taking or avoiding tolled road are as follows:

Company location has a positive effect to the utility of using toll roads. If the company is located in Bangkok or in any of its vicinities, they might be likely to take toll roads. The reason is that they want to avoid traffic congestion in Bangkok area and they want to improve their performance service in terms of time saving and time reliability.

Type of goods carried is positively effect by the utility of using toll roads. If the company carries perishable goods or products, such as food, concert etc., they might take toll roads. They have to use certain and reliable routes. Because their goods or product may be perishable.

The types of facilities the most frequently transport goods has a positive effect on the utility of using toll roads. If the company carries goods to or from an airport, they might take toll roads, because their products have high value, thus they should take routes which are certain and reliable.

The number of truck drivers is negatively affect by the utility of using toll roads. If the company has a high number of truck drivers, they are indisposed to taking toll road.

The number of trucks in a fleet is negative effect by the utility of using toll roads. If the company has a high number of trucks, they are indisposed to taking toll road.

The typical operating environment is positively effect by the utility of using toll roads. If the company carries goods to or from rural area, they might take toll road, because they are transporting over long distances and they need to improve their performance service in terms of time savings and time reliability.

The toll payer has a positive effect on the utility of using toll roads. If the company pays the tolls, the drivers are usually willing to take toll road.

Truck loads positively affect the utility of using toll road. If the company carries goods with truck load (full), they might take toll road.

Distance positively affects the utility of using toll roads. If the company carries goods over long routes, they might take toll road.

Carrier time has a positive effect on the utility for using toll roads. If the company carries goods during peak times of day (morning peak, afternoon peak), they might take toll roads. They do so because they prefer to avoid traffic congest in the Bangkok area and improve their performance service in terms of time saving and time reliability.

4.2 Web Survey Design

The research was collected from the questionnaire survey, with choice experiments, by using a website. The advantage of a website survey is high efficiency in managed data and reduced data bias. The flowchart of the website survey is shown in Figure 4.4.

Part 1st

Respondent characteristics and company background

Part 2nd

Choice game

Auto input choice experiment

Random choice experiment

Choice remaining

Get answer

Program ASP

No choice remaining

End part 2nd

Part 3rd

Route choice decision

End

Figure 4.4 Flowchart of webbase survey

The website was developed by using PHP scripting language, HTML 5 scripting language and MySQL database. The three main website interfaces and two minor interfaces are as follows: 1) Part 1st Respondent characteristics and company background, 2) Part 2nd Choice game, 3) Part 3rd Route choice decision, 4) Welcome interface and 5) End interface.

4.2.1 Logistic Manager and Firm Characteristics

The flowchart of this part was shown in Figure 4.5. The purpose of the first section was to get general information about respondents and their business. Questions were asked about their: age, gender, education, job position, company’s location, type of goods carrier and business classification. The flowchart of this part was shown in Figure 4.5.

The 1st step

Respondent characteristics and company background form

The 2nd step

The respondent fills the form and input their data

Form not complete

Form complete

The 3rd step

The data is recorded into "respondent table" in database

End

Figure 4.5 Flowchart of the 1st Part

The interface of the first part was shown in Figure 4.6. When the respondents come to this part, the interface will show the form: age (fill), gender (check), education (fill), job position (fill), your company location (fill), type of goods carrier (check) and business classification (check). Then, the respondents have to complete the form. If the form is not complete, the interface will come back to step 2nd. When all of the form is completed, the data is recorded to the table, "respondent", in the database. Then, the website will go to the next part.

Figure 4.6 1st part interface

4.2.2 Choice Game

A Choice Game is Stated Preference experiment that consists of choices between alternatives described by attributes set to particular levels. Adaptive Stated Preference (ASP) makes it possible to enter responses at the time of the interview. Most importantly, ASP reduces the possibilities for mistakes by using algorithms. ASP takes the process one step further, and amends attribute levels offered at later stages of the experiment in light of responses to earlier stages. For example, a respondent who would not pay ฿200 for a new transport facility, the survey would not ask if they would pay ฿100.

Choice game interface is adaptive SP data collection that was shown in the interface in Figure 4.7. The interface uses a three row format, with row 1 (toll cost), row 2 (travel time) and row 3 (travel time reliability).

There are twenty six choice experiments in the total. For each experiment was randomly presented on a computer screen. When the users filled out the form, the website processed by using the algorithms until no choice remained. The interface of the second part is shown in Figure 4.8.

Recorded into "random" table in database

The 1st step

Auto input choice experiments

The 2nd step

Random choice experiment from table "random"

The 3rd step

The respondent fills the form and inputs their data

Recorded into "game" table in database

Record into "game" table in database

The 4rd step

ASP processing by using algorithm

Remain choice in "random" table

No choice in "random" table

End

** Note "game" table record answers of each choice experiment

"random" table record choice experiment that are remain

Figure 4.7 Flowchart of the 2nd Part

The algorithm was analyzed by the response answer; For example, there are 4 choices with travel cost ฿100, ฿200, ฿300 and ฿400 for using a new facility. If a respondent would pay ฿200 for new transport facility, they would pay ฿100 to take the facility. Therefore, there are ฿300 and ฿400 in the random choice set. The process will loop until there are no choices in the choice set and all answers have been received. The algorithm of the ASP programming shows in appendix C.

Figure 4.8 2nd part interface

4.2.3 Route Choice Decision

The purpose of third part is to understand how businesses, when moving goods, assess tradeoffs in using or avoiding toll facilities. The question asked is shown in Appendix D. The flowchart of this part is shown in Figure 4.10.

The 1st step

Route choice decision form

The 2nd step

The respondent fills the form and inputs their data

Form not complete

Form complete

The 3rd step

The data is recorded into "firm table" in database

End

Figure 4.9 Flowchart of the 3rd Part