III. Research Methodology
3.4. Survey Questionnaires
For this research survey, mySurvey29 is used for the confection of the questionnaires survey due that this website provide tool to develop AHP questionnaire. Following, some sample illustrating the survey questionnaire format. Basic information of participant was required such as respondent profession area and experience in the area.
Figure 30. Respondent’s basic information
Source: This research data
In figure 30 shows how respondent’s basic information format look like. Then, as AHP is used to make pair wise comparison among criteria, Saaty scale format is used. Saaty scale have 9 degree intensity such as Absolute important represented by 9, Very Important represented by 7, Essential Important represented by 5,Weak Important represented by 3, Equal Important represented by 1 and the number 2, 4, 6, 8 are middles value. In figure 31 shows how the format for an AHP survey questionnaire is.
29 Website: www.mysurvey.tw
76
Subsequently, the elaboration of the pair wise comparison matrix is done by layer, which means that first pair wise comparison matrixes is done for criteria, and then follow to the lower layer with pair wise comparison matrix. In figure 31 shows on the AHP Framework how many layers have.
Figure 31. Questionnaire Survey Format
Criteria A
Absolute important Very Important Essential Important Weak Important Equal Important Weak Important Essential Important Very Important Absolute important
Criteria B
9 7 5 3 1 3 5 7 9
C1 C2
C1 C3
C2 C3
Source: This research data
Figure 32.Thesis Research AHP Framework
Source: This research data
77
Figure 33. Sample of Questionnaire Survey Format for Criteria
Source: This research data
In figure 33 shows a sample of Questionnaire Survey Format for Criteria while in figure 34 shows a sample of Questionnaire Survey Format for Finance sub-criteria
Figure 34. Sample of Questionnaire Survey Format for Finance sub-criteria
Source: This research data
78 3.5. Analytical Hierarchy Process Model
Analytical Hierarchy Process model is based on four steps: problem modeling, weights valuation, weights aggregation and sensitivity analysis.30
3.5.1. Problem modeling
In a decision-making processes, the structure the problem, which can be divided into three parts: goal (Stakeholder satisfaction), criteria (criteria and its sub-criteria) and alternatives (in this research no alternative is presented). AHP has the advantage of permitting a hierarchical structure of the criteria, which provides users with a better focus on specific criteria and sub-criteria when allocating the weights.
The first step in the AHP model is to define the goal and objectives of the research.
Goal and objectives set up as shown in table 31. Goal and objectives. This step is important, because a different structure may lead to a different final ranking.
Table 31. Goal and objectives Goal and Objectives
Goal Satisfaction of stakeholder
Objectives 1. Improve the mobility of people and goods in the region.
2. Promote economic prosperity.
3. Protect the natural environment.
4. Promote an overall high quality of life.
5. Distribute transportation benefits and costs equitably.
6. Create an efficient land use pattern for the provision of infrastructure, facilities, and services.
Source: This research data
Following, model the structure in a hierarchy way containing the goal as the higher layer of the structure. Subsequent, follows by criteria as the second layer and the sub-criteria
30 Alessio Ishizaka and Ashraf Labib, Analytic Hierarchy Process and Expert Choice: Benefits and Limitations, 22(4), p. 201–220, 2009.
79
as the third layer of the hierarchical structure as shown in figure 35. Analytic Hierarchy Process framework,
Figure 35. Analytic Hierarchy Process framework
Source: This research data 3.5.2. Pairwise Comparisons
Figure 36. Pairwise comparison example
Criteria C1 C2 C3 C4 C5 C6 C7 Sub-criteria M1 M2 M3 M4
C1 1 M1 1
C2 1 M2 1
C3 1 M3 1
C4 1 M4 1
C5 1
C6 1
C7 1
Source: This research data
At each node of the hierarchy, a matrix will collect the pairwise comparisons of the decision-maker. Then, establish matrices of pair-wise comparisons according to AHP framework and questionnaire results. As this study AHP has 3 layer levels, where the second level is located the criteria and the last layer levels located each criterion sub-criteria. Then, as there are 7 criteria approximate 8 matrices of pair-wise comparisons are established.
80 3.5.3. Judgment Scales
Table 32. The AHP verbal scale ranges 1 to 9 (Saaty scale) Intensity of
importance Definition Explanation
1 Equal importance Two elements contribute equally to the objective
3 Weak importance Experience and judgment slightly favor one element over another
5 Essential importance Experience and judgment strongly favor one element over another
7 Very importance One element is favored very strongly over another; its dominance is demonstrated in practice
9 Absolute importance Extreme importance
Intensities of 2, 4, 6, and 8 can be used to express intermediate values.
Source: Adapted from Saaty
One of AHP’s strengths is the possibility to evaluate quantitative as well as qualitative criteria and alternatives on the same preference scale of nine levels. These can be numerical, verbal or graphical (Expert Choice).
3.5.4 Priorities derivation
Once the comparisons matrices are filled, priorities can be calculated. The traditional AHP uses the eigenvalue method. Computation of eigenvalue by the relative weights the criteria and the sum is taken over all weighted eigenvector entries.
Pair wise comparison data can be analyzed using the eigenvalue technique. Using these pair wise comparisons, the parameters can be estimated. The right eigenvector of the largest eigenvalue of matrix A constitutes the estimation of relative importance of attributes.
Then, if the matrix A is consistent, then A contains no errors (the weights are already known) and we have:
81
The which in matrix notation, is equivalent to Aw=nw. The vector w is the principal right eigenvector of matrix A corresponding to the eigenvalue n. If the vector of weights is not known, then it can be estimated from the pair wise comparison of matrix A generated by the decision maker by solving for:
The matrix A contains the pair wise judgments of decision maker and approximates matrix A whose entries are unknown.
3.5.5. Consistency
As priorities make sense only if derived from consistent or near consistent matrices, a consistency check must be applied. Several other methods have been proposed to measure consistency. Peláez and Lamata (2003) describe a method based on the determinant of the matrix. Crawford and Williams (1985) prefer to sum the difference between the ratio of the calculated priorities and the given comparisons. The transitivity rule has been used by Salo and Hamalainen (1997) and later by Ji and Jiang (2003). Alonso and Lamata (2006) have computed a regression of the random indices and propose the formulation: λmax < n + 0.1(1.7699n-4.3513). Stein and Mizzi (2007) use the normalized column of the comparison matrix.
Saaty (1977) has proposed a consistency index (CI). Saaty (1977) has shown that the largest eigenvalue, l max of a reciprocal matrix A is always greater than or equal to n. If the pair wise comparisons do not include any inconsistencies, lmax = n. The more consistent the maximum comparisons are, the closer the value of computed lmax to n.
A consistency index (CI), which measures the inconsistencies of pair wise comparisons is given in:
82
Where CI is the consistency index; λmax is the maximal eigenvalue; RI is the random index; and n is the number of columns. The RI is the average of the CI of a large number of randomly generated matrices, where n is the matrix size. Judgment consistency can be checked by taking the CR of CI with the appropriate value in Table 33. If CR is less than 10%, then the matrix can be considered as having an acceptable consistency (Saaty, 1980).
Table 33. Consistency Table
Size of
matrix 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Random
Consistency 0.00 0.00 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 1.51 1.48 1.56 1.57 1.58
Source: Adapted from Saaty
3.5.6. Aggregation
The last step is to synthesize the local priorities across all criteria in order to determine the global priority. The historical AHP approach (called later distributive mode) adopts an additive aggregation with normalization of the sum of the local priorities to unity
Where pi is the global priority of the alternative i; lij is the local priority; wj is the weight of the criterion j.
3.5.7. Geometric Mean
In case that researcher doesn’t have the software Expert Choice, there is other way to compute an overall result; researcher could make the evaluation of ranking intensity. The geometric mean is used in the AHP model for the overall assessment in case that there have more than 1 respondent. The formula of geometric mean followed as follow:
83 3.5.8. Sensitivity Analysis
The last step of the decision process is the sensitivity analysis, where the input data are slightly modified in order to observe the impact on the results. If the ranking does not change, the results are said to be robust.
3.5.9. Expert Choice
As the Analytic Hierarchy Process (AHP) as it is implemented in the software package Expert Choice. The software Expert Choice31 is available on the market which simplifies the execution of the AHP process and automates many of its computations. In short, there is no need to implement the steps manually. However, manual computation can be done with help of Microsoft Excel. Microsoft Excel can help to ease the computation process.
31 www.expertchoice.com
84
IV. Data Analysis
4. Data Processing and Analysis Method
As the model used for this research is Analytical Hierarchy Process, the structure of questionnaire use the Saaty scale; absolute important, very important, essential important, weak important, equal important that are represented by 9, 7, 5, 3, and 1, respectively.
The AHP structure have 3 layer; goal, criteria and sub-criteria. We have 7 major criteria such as (1) mobility, (2) finance, (3) growth management, (4) economic prosperity, (5) environmental stewardship, (6) quality of life, and (7) equity.
This study has a total of 27 sub-criteria which are distributed in the following way: (1) mobility has 4 criteria, (2) finance has 4 criteria, (3) growth management has 4 criteria, (4) economic prosperity has 3 criteria, (5) environmental stewardship has 5 sub-criteria, (6) quality of life has 3 sub-sub-criteria, and (7) equity has 4 sub-criteria.
4.1. Data Collection
Table 34. Data collection-Stakeholder request
Survey Data Collection
High -Tech Industries Airlines Industries Government Total
Request Questionnaire 10 10 10 30
Response Questionnaire 9 7 10 26
Valid Questionnaire 7 6 8 21
Invalid Questionnaire 2 1 2 5
Valid Questionnaire rate 70%
Source: This research data
A total of 30 questionnaires were mail to the different stakeholder from May 1, 2012, the questionnaire recovery as shown in the table 28 is presented the request for survey for the different stakeholder. Only 26 people reply back the questionnaire where 21 are valid
85
questionnaire and the others 5 are invalid questionnaire. Then, the average of valid questionnaire rate is about 70%.
4.2. Test and Weighting Computation
Table 35. Criteria Inconsistency
Second Layer Third Layer
Major Criteria Mobility Finance Growth M Economic P Environmental Quality of Life Equity
CI CI CI CI CI CI CI CI
Firstly, need to check if the consistency test of the questionnaire (CI) is good enough, and seek consistency at all levels of the proportion of eigenvalues (at λmax). Knife root
86
smoked (1993) pointed out that the CI and CR, preferably less than 0.1, as the case may sometimes be allowed to 0.15. However, Deng Zhenyuan (2002) pointed out that the maximum allowable range of the CI value of 0.2. As in this research study most of the data consistency is around 0.2 then data selected for the study would have a CI and CR range below 0.2. From the 26 questionnaire collected, 5 questionnaire are invalid due that their CI and CR are higher than 0.2 as shown in the table below. The invalid questionnaire corresponds to participant 1, 3, 5, 10, and 17.
4.3. Analysis of all the AHP’s level
Each stakeholder has a different opinion and need about their business industries therefore, a variety of stakeholder respondents were needed to understand the important of the criteria. As there are more than 1 participant, a geometric mean is used for the computation of all the matrix of the AHP’s level and weighting as shown in the below tables.
4.3.1. Major criteria
The impact of the Cross Strait Air Policy on aviation industries take in consideration 7 criteria such as (1) mobility, (2) finance, (3) growth management, (4) economic prosperity, (5) environmental stewardship, (6) quality of life, and (7) equity.
Table 36. Airlines Industry - Major Criteria’s Weighting
M F GM EP ES QL E Weighting Ranking
M 1 1/1.99 1/2.61 1/1.94 1.20 1.73 1.27 0.105 5
F 1.99 1 1.44 1.73 3.13 4.09 2.39 0.263 1
GM 2.61 1/1.44 1 1 1.05 3.00 2.75 0.192 2
EP 1.94 1/1.73 1 1 1.66 3.31 2.50 0.187 3
ES 1/1.20 1/3.13 1/1.05 1/1.66 1 1.10 1.57 0.108 4
QL 1/1.73 1/4.09 1/3.00 1/3.31 1/1.10 1 1.10 0.069 7 E 1/1.27 1/2.39 1/2.75 1/2.50 1/1.57 1/1.10 1 0.076 6 Lambda: 4.045, C.I.:0.02, C.R.:0.02
M: mobility, F: finance, GM: growth management, EP: economic prosperity, ES:
environmental stewardship, QL: quality of life, E: equity.
Source: This research data
As shown in the table 36, Airlines Industry care more about finance, growth management, economic prosperity, environmental stewardship, mobility, equity and quality of
87
life. This industry weights these criteria in the following way: 0.263, 0.192, 0.187, 0.108, 0.105, 0.076 and 0.069, respectively. While this level of the consistency test is CI = 0.02, CR
= 0.02, are less than 0.1, then CI and CR fill with the requirements of consistency.
Table 37. High Tech Industry – Major Criteria’s Weighting
M F GM EP ES QL E Weighting Ranking stewardship, QL: quality of life, E: equity.
Source: This research data
On the other hand, High Tech industry concern more about finance, mobility, environmental stewardship, growth management, equity, quality of life and followed by economic prosperity. In contrast, as shown in table 38, Government considers as priority environmental stewardship, quality of life, equity, economic prosperity, growth management, finance and then mobility. All these results are useful due that the CI and CR of both table fix in with the consistency value range established.
Table 38. Government – Major Criteria’s weighting
M F GM EP ES QL E Weighting Ranking stewardship, QL: quality of life, E: equity.
Source: This research data
88
Table 39. Overall Assessment- Major Criteria’s Weighting
M F GM EP ES QL E Weighting Ranking Lambda: 3.099, C.I.:0.00761, C.R.:0.00761
M: mobility, F: finance, GM: growth management, EP: economic prosperity, ES:
environmental stewardship, QL: quality of life, E: equity.
Source: This research data
And finally, at table 39, we have the assessment of the 3 stakeholder together. As shown in the table, in an overall point of view, the stakeholder would concern more about finance, growth management, environmental stewardship, economic prosperity, quality of life, equity and mobility.
4.3.2. Mobility
With the mobility factor, we wanted to define better which benefits gain is more appreciated by users of the Cross Strait direct flight. As same as above, we have separate assessment by stakeholder and ending with a group assessment. The purpose is to see the opinion of each industry in an individual point of view and then, in a group point of view.
In the airlines industry, respondents take more emphasis in the “travel time saving”
with a weighting level of 0.319. Following the travel time saving, respondents desire high reliability, vehicle operating and ownership and the last one, other user benefits.
Table 40. Airlines Industry – Mobility Weighting
M1 M2 M3 M4 Weighting Ranking
M1: travel time saving, M2:reliability, M3:vehicle operating and ownership, M4:
other user benefit
Source: This research data
89
Stakeholder strongly care about time saving especially in for airlines industries and technology industries due to the behavior and characteristics of their business, they require certain level of tight logistic performance. For airlines industries, saving in travel time represent a benefit for the wages paid to their labor especially pilot and air hostess. Also, shorten flight distance can represent less consumption of fuel. As shown in the figure 37.
Mainland China and Taiwan airlines expenses, labor and fuel cost represent 48% above of their operating cost. Past studies suggested that lower cost in freight movement have a positive effect on all firms engaged in production, distribution, trade or retail sale of physical goods.
Figure 37. Mainland China (left side) and Taiwan (right side) airlines expenses
Source: Airlines Annual Report (2010)
On the other hand, high tech industry respondent appreciated more the reliability with a weighting of 0.419. Next, this industry care about the mobility, vehicle operating and ownership and other user benefits. High tech industry weights these criteria as follow 0.243, 0.186, and 0.154, respectively.
As many literature mentioned, users of transportation system value travel time reliability as well as saving since greater predictability in travel time help to reduce the cost associated with activity scheduling. Greater schedule reliability will have significant impacts in term of time gains. Therefore, allow high tech firms to manage their inventories and supply chains more efficiently. Increased in reliability reduce the requirement for buffer stock, inventory held to protect against delivery failure.
90
Table 41. High Tech Industry – Mobility Weighting
M1 M2 M3 M4 Weighting Ranking
M1: travel time saving, M2:reliability, M3:vehicle operating and ownership, M4:
other user benefit
Source: This research data However, government considers more desire improving the “travel time saving” which weight it with 0.380. Then followed by the criteria “reliability” weighting it with 0.324;
“vehicle operating and ownership”, weight it with 0.185. And last, government considers the criteria “other user benefits”, the level of a weight of 0.111.
Table 42. Government - Mobility Weighting
M1 M2 M3 M4 Weighting Ranking
M1: travel time saving, M2:reliability, M3:vehicle operating and ownership, M4: other user benefit
Source: This research data Regarding operation and ownership of vehicles, it involves real resource costs that are associated with using the transportation system. There are extensive literature that related vehicle operating and ownership with the changes in industries networks. For airlines is essential the selection of an airport point where their hub would be located. This means that after the Cross Strait direct air policy, carriers can planned better their hub and create better schedule of their fleet, employer and routing. Carriers that can provide better service can influence shipper business and logistics performance.
At last, the table 43 shows the overall assessment of the stakeholder’s point of view for the criteria mobility. In an overall assessment, stakeholders appreciate the gain of mobility
91
benefits in the following order: reliability, travel time saving, vehicle operating and ownership, and other user benefit.
Table 43. Overall Assessment- Mobility Weighting
M1 M2 M3 M4 Weighting Ranking
M1 1 1.03 1.74 1.74 0.318 2
M2 1/1.03 1 2.06 2.36 0.345 1
M3 1/1.74 1/2.06 1 1.68 0.195 3
M4 1/1.74 1/2.36 1/1.68 1 0.142 4
Lambda: 4.452, C.I.:0.01, C.R.:0.01
M1: travel time saving, M2:reliability, M3:vehicle operating and ownership, M4:
other user benefit
Source: This research data
4.3.3. Finance
As shown in the tables below, with the criteria finance, we try to capture what cost is more concerning for transportation users. In the case of airlines industry, they concern more about capital cost, followed by operating revenue, then, operating cost and finally, influence of finance on economy. However, high tech industry would concern more about operating revenue, operating cost, capital cost and at last, influence of finance on economy.
Figure 38. Mainland China (left side) and Taiwan (right side) airlines revenue
Source: Airlines Annual Report (2008)
The criteria “operating revenue” is relevant for this kind of industry due to the nature of their business. As carriers measure their profitability by the seat and weight load factor, it is very important for them to know what make their business. As shown in the figure 38.
92
Mainland China (left side) and Taiwan (right side) airlines revenue, passenger activity represents more than 90% of the total revenue gain by Mainland China airlines. On the other hand, Taiwanese airline’s revenue represents almost half of cargo and half of passenger activities. However, high tech industries rely more in transportation to improve their logistic activities which allows the industries to reduce their cost and allow them to serve a wider market.
Besides operating revenue, the operating cost is another big concern especially for transportation industry as they rely heavily in petroleum and qualified professional. As mentioned in figure 37. Mainland China (left side) and Taiwan (right side) airlines expenses, labor and fuel are very important for airlines. On the other hand, many Taiwanese manufacturing established their factories in Mainland China are due to lower labor cost.
Table 44. Airlines Industry – Finance Weighting
F1 F2 F3 F4 Weighting Ranking
F1: operating cost, F2:capital cost, F3:operating revenue, F4: influence of finance on economy
Source: This research data
Table 45. High Tech Industry – Finance Weighting
F1 F2 F3 F4 Weighting Ranking
F1: operating cost, F2:capital cost, F3:operating revenue, F4: influence of finance on economy
Source: This research data
On the other hand, government would concern about operating revenue, influence of finance on economy, operating cost, and capital cost. Based on government’s respondent, they
93
weight the criteria “operating revenue” with 0.354. Subsequently, governments weight the criteria “influence of finance on economy” with 0.286. Next, the criteria “operating cost” with 0.197, and capital cost, weight it with 0.163.
Table 46. Government – Finance Weighting
F1 F2 F3 F4 Weighting Ranking
F1: operating cost, F2:capital cost, F3:operating revenue, F4: influence of finance on economy
Source: This research data
Table 47. Overall Assessment – Finance Weighting
F1 F2 F3 F4 Weighting Ranking
F1: operating cost, F2:capital cost, F3:operating revenue, F4: influence of finance on economy
F1: operating cost, F2:capital cost, F3:operating revenue, F4: influence of finance on economy