Hospitality Management 21 (2002) 171187

Multicriteria selection for a restaurant location

in Taipei
Gwo-Hshiung Tzeng
a,
*, Mei-Hwa Teng
a
, June-Jye Chen
b
,
Seram Opricovic
c,1
a
Institute of Trafc and Transportation, and Institute of Information Management,
National Chiao Tung University, 114, 4F, Sec. 1, Chung Hsiao W. Road, Taipei 100, Taiwan
b
Department of Hotel Management, Ming-Shin Institute of Technology, 1 Hsin-Hsing Road,
Hsin-Fong, Hsin-Chu, Taiwan
c
Faculty of Civil Engineering, University of Belgrade, Belgrade, Yugoslavia
Abstract
One of the most important factors leading to the success of a restaurant is its location. A
multicriteria decision-making method is used to rank alternative restaurant locations. The set
of criteria is established for the Taipei case, and corresponding criteria should be established
for each case study, although this multicriteria decision-making approach has broader
applicability. In this paper, the analytic hierarchy process (AHP) with ve aspects and 11
criteria is used to develop a location evaluation hierarchy for a restaurant. Four alternatives
for the Pao-San (Takarazima Japanese Siki Kisegi cuisine) restaurant location in Taipei are
evaluated. The aspects include transportation, commercial area, economic, competition and
environment. The criteria are rent cost, transportation cost, convenience to mass transporta-
tion system, size of parking space, pedestrian volume, number of competitors, the intensity of
competition, size of the commercial area where the restaurant is located, extent of public
facilities, convenience of garbage disposal, and sewage capacity. The alternatives are ranked by
a multicriteria method. The result is the set of compromise solutions, including two alternative
locations, to be proposed to the decision maker. The rst alternative is in an expanding
commercial center at the intersection of two subway lines. The second alternative is in a new
city political and administrative center. r 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Restaurant location; Multicriteria decision-making; Compromise solution
*Corresponding author. Tel.: +886-3-5712121; fax: +886-3-5753926.
E-mail addresses: ghtzeng@cc.nctu.edu.tw (G.-H. Tzeng), jjchen1@ms32.hinet.net (J.-J. Chen),
seropric@yahoo.com (S. Opricovic).
1
Visiting professor at National Chiao Tung University, supported by National Science Council of
Taiwan (1999/2000), Taiwan.
0278-4319/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 2 7 8 - 4 3 1 9 ( 0 2 ) 0 0 0 0 5 - 1
1. Introduction
Restaurant location is an important factor leading to the success of restaurant
management since it will determine the convenience of service to customers and how
many are attracted. In addition, location also exerts a signicant effect on market
share and prot-making ability of the restaurant. Jain and Mahajan (1979) pointed
out that, in the development of competition strategy, price can help to cope with,
service can help to expand and improve, commodity can be overlapping, promotion
can be imitated, however, the benets from the establishment location to the retailer
could hardly be undermined and undervalued. Therefore, the importance of the
location should be appreciated.
A restaurant is a project that provides for diners, as well as a place where the
general public conducts parties and other social events. Thus, restaurants should
focus both on providing service to individual customers (Yun and Hing, 1995), and
also on the required service characteristics to the customers as a whole. The
categories of service such as location, size, convenience of transportation, and
parking capacity, will directly affect the sources of customers and the future success
of the restaurant. Therefore, the selection of restaurant location becomes more
important.
There are a number of papers discussing the store location problem. Nelson (1958)
discussed the principles of retail location selection. Applebaum and Cohen (1960)
evaluated store sites and determined store rents. Kem and Simmons (1990) proposed
criteria for location selection. Jain and Mahajan (1979) evaluated the competitive
environment in retailing by using multiplicative competitive interactive models. In
Yun and Hing (1995), an instrument for measuring quality in restaurant operations
is applied. In Olsen et al. (1998), strategic management in the hospitality industry is
considered, including managing service quality, and the broader aspects of
globalization, multinationals, and corporate strategies.
Within the restaurant location selection process, evaluation has to be both from
the perspective of customers and from management. The lifestyles and convenience
for customers have to be considered, as well as the cost and management problems of
the restaurateur. This paper has, therefore, applied multiple criteria decision making
(MCDM) to evaluate restaurant location alternatives, determining the importance
(weight) of each evaluation criterion. As an illustrative example, the restaurant
location problem for the Takarazima Japanese Siki Kisegi cuisine restaurant is
solved.
2. The evaluation model for selection of the restaurant location
The restaurant location selection is exposed to the inuences of two types of
factors: external and internal. External factors include economic, transportation,
competition, commercial area and environment, while internal factors include the
personal background of the manager (such as age), ideas of management, level of
education, decision-making preference, and management system. The internal
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 172
factors, which depend on the attributes of the restaurant management (Guerrier and
Deery, 1998), are not included in this model.
In the decision-making process, representatives of the catering business, scholars,
and experts are involved. The hierarchical structure of evaluation has been construed
according to all aspects. The criteria were established through the analytic hierarchy
process (AHP) (Saaty, 1980).
The evaluation principles for location selection are established, with the help of
scholars, caterers, and expertise of the restaurateurs. The hierarchical structure of
evaluation is presented in Table 1. The rst level is the goal level, and the ultimate
goal of evaluation for the location selection of the restaurant is to determine the
optimal location. The second level is the objective level, which includes economic,
transportation, competition, commercial area, and environment factors. The third
level is the attribute level, which includes 11 evaluation criteria.
When investors invest in a restaurant, they want to evaluate costs and benets.
The costs, which are part of the economic aspect, should include personnel, material,
rent, and transportation costs. In the location selection model, only rent and
transportation costs are included, since personnel and material costs are not
signicant in the location selection problem. Rent cost denotes the monthly amount
of money needed for the land and building used for the restaurant. Transportation
cost refers to the monthly cost to deliver the materials and goods purchased for
operating the restaurant.
In the location selection problem it is also necessary to take the transportation
convenience of customers into consideration. Three separate criteria are considered
in this transportation aspect: convenience to mass transportation, parking capacity,
and pedestrian volume. Convenience to mass transportation refers to the number of
bus routes close to the restaurant, more precisely the number of bus routes within
Table 1
Evaluation hierarchical structure for restaurant location selection
Goal Objectives Criteria
Optimal location of restaurant 1. Economic 1. Rent cost
2. Transportation cost
2. Transportation 3. Convenience of mass transportation system
4. Parking capacity
5. Pedestrian volume
3. Competition 6. Number of competitors
7. Intensity of competition
4. Commercial area 8. Size of commercial area
9. Extent of public facilities
5. Environment 10. Convenience of garbage disposal
11. Sewage capacity
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 173
500 m around the restaurant, with more bus routes indicating greater convenience.
Restaurant customers are also likely to drive their own cars, so more parking spaces
can attract more customers to dine in the restaurant. The parking capacity, as a
criterion, is expressed as the number of parking units within 500 m around the
restaurant, including those parking lots related to the restaurant, public parking and
private parking lots. The people passing by the restaurant are also potential
customers of the restaurant, so the pedestrian volume is very important to the
restaurant. The pedestrian volume refers to the number of pedestrians passing by the
restaurant when it is open.
The number of competitors and intensity of competition are included as two
criteria. The number of competitors refers to the number of similar restaurants in the
vicinity. For example, low-priced seafood restaurants in the vicinity are similar
restaurants in this case study. The intensity of competitors refers to the scale of
restaurant in the vicinity. The scale of restaurant includes management features,
capacity of the restaurant, atmosphere of restaurant, and specialty of cuisine,
because all these items can be taken as competitive edge. Larger scale represents
stronger competition from other restaurants.
The commercial area aspect includes two criteria: the size of the commercial area
and the extent of completeness of public facilities. The size of the commercial area
where the restaurant is located denotes the geographical extent of the commercial
area, with greater range indicating wider commercial area. Thus, if the restaurant is
located in a large commercial area, then there will be greater potential benet. The
extent of public facilities refers to the following public facilities: transit railway
station, bus stop, community center, art and cultural center, stadium, gymnasium
and public library. Easier access and more complete facilities mean that more
customers will come to the restaurant.
The environmental objective includes two separate criteria: convenience of
garbage disposal and management competence of sewage. The convenience of
garbage disposal refers to the easiness of handling garbage produced daily by the
restaurant. Therefore, the distance to the garbage collection center will affect the
convenience. Sewage capacity denotes whether the local sewage system can
adequately and suitably handle the sewage daily produced by the restaurant without
impacting the neighboring environment.
3. Assessment of criteria weights
Fifteen evaluators, including ve scholars, ve catering experts and ve
restaurateurs, were invited to evaluate the weights of the objectives and criteria.
The names of experts were chosen with the assistance of the general manager of
Takarazima restaurant. Data were collected interviewing experts. An interview was
conducted with a questionnaire, and answers were expressed in the form of a
pairwise comparison matrix. This weights assessment procedure is based on the AHP
(Saaty, 1980; Saaty and Vargas, 1982; Appendix A), starting with hierarchy in Table
1, analyzing data from the experts, and ending with weights in Tables 2 and 3. The
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 174
results are shown in Tables 2 and 3. The coefcient of variation s= % x indicates
the consensus of the group of evaluators; smaller coefcients indicate higher
consensus.
The ranking by objective weights (Total in Table 2) is as follows: transporta-
tion, commercial area, economic, competition and environment. Transportation is
the most important objective for several reasons. Generally, the trafc condition in
Taipei City is quite bad, so trafc jams are frequent, and people cannot easily nd
parking spaces (ITT Project, 1998; THI Study, 1996). Although there are two rapid
transit routes under operation, the subway system is still under construction, and the
bus system is most frequently used for public transportation.
Environment is considered as a less important objective in this evaluation. The
main sewage pipes in the Taipei sewage system have already been set up, and the
restaurants can easily connect to the sewage pipes. In addition, garbage is removed
three times per week and the density of garbage collection points is quite satisfactory,
so restaurants can easily handle the garbage produced.
Table 2
Weights (w) of the objectives and coefcients of variation (CV)
Objectives Scholars Experts Restaurateurs Total
w r CV w r CV w r CV w r CV
Economic 0.2603 1 0.5736 0.2090 3 0.3827 0.2239 3 0.4851 0.2311 3 0.4541
Transportation 0.2351 3 0.4167 0.2662 1 0.4115 0.2777 1 0.4245 0.2597 1 0.3800
Competition 0.1767 4 0.2263 0.1790 4 0.5299 0.1265 4 0.2235 0.1607 4 0.3226
Commercial area 0.2361 2 0.6790 0.2488 2 0.4272 0.2759 2 0.4424 0.2536 2 0.4753
Environment 0.0919 5 0.5548 0.0970 5 0.3717 0.0960 5 0.3125 0.0949 5 0.4075
Note: r denotes the rank of objective according to its weight; CV=standard deviation (s)/mean % x:
Table 3
Criteria weights (w) and coefcients of variation CV
Criteria Scholars Experts Restaurateurs Total
w CV w CV w CV w CV
Rent cost 0.2006 0.6613 0.1698 0.3906 0.1721 0.5722 0.1808 0.5219
Transportation cost 0.0596 0.6049 0.0393 0.3598 0.0518 0.5107 0.0502 0.4966
Conv. of mass transport. 0.0954 0.6021 0.0639 0.6259 0.0622 0.6821 0.0739 0.5296
Parking capacity 0.0735 0.9327 0.0848 0.7998 0.1154 0.6367 0.0912 0.6873
Pedestrian volume 0.0661 0.6051 0.1175 0.6481 0.1000 0.6164 0.0945 0.5576
Number of competitors 0.1102 0.4537 0.1237 0.5773 0.0740 0.5233 0.1026 0.4563
Intensity of competitors 0.0665 0.5626 0.0553 0.5718 0.0525 0.3299 0.0581 0.4657
Size of commercial area 0.1782 0.7186 0.1714 0.4326 0.1843 0.4730 0.1779 0.5187
Extent of public facility 0.0580 0.6216 0.0774 0.6836 0.0916 0.5002 0.0756 0.5454
Conv. of garbage disposal 0.0472 0.4237 0.0535 0.6993 0.0469 0.5641 0.0492 0.5394
Sewage capacity 0.0447 0.8948 0.0435 0.2298 0.0491 0.4988 0.0457 0.6139
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 175
The objective weights of an individual group indicate that there exist differences of
opinion towards each of the objectives. The scholars feel that economic criteria are
most important, while catering experts and restaurateurs place greatest importance
on transportation (Table 2).
The consensus can be justied by the coefcient of variation of the weights
for each of the objectives, with lower coefcients indicating higher consensus.
The Total values in Table 2 show that the highest consensus of all evaluators is
for competition, while the lowest consensus is for commercial area. The scholars
have the highest consensus for the competition, while the lowest is for commercial
area; the catering experts have the highest consensus for environment, and
the lowest for competition; the restaurateurs have the highest consensus for
competition, and the lowest for economic. The scholars and restaurateurs are quite
consistent in their opinions about competition, while the catering expert group tends
to be most consistent in terms of environment, but has a larger disparity in
competition.
Table 3 presents the weights for the 11 criteria. The data indicate that the rent cost
and the size of the commercial area are more important criteria, while the
convenience of garbage disposal and management ability of sewage are less
important criteria.
The criteria weights given by individual groups indicate that there is a mild
disparity of opinion for each criterion within each group. Scholars consider rent cost,
size of commercial area where the restaurant is located, and the number of
competitors to be of greatest importance. Catering experts observe the size of
commercial area where the restaurant is located, rent cost, and number of
competitors as more important. Restaurateurs consider the size of commercial area,
rent cost, and size of parking space, as the most important.
Consensus on the criteria can be determined by the coefcient of variation of the
weights for each criterion, as presented in Table 3. The consensus is quite low
compared with the consensus on the objectives because most of the coefcients of
variation are greater than the coefcients in Table 2. It seems that the importance of
the criteria is quite different in the opinion of the evaluators.
As a relative measure of variability we used the coefcient of variation (CV=the
standard deviation divided by the mean). But it should be pointed out that the
coefcient of variation loses its utility as a measure of variability when the mean
value is close to zero since small changes in the standard deviation would cause large
changes in the coefcient of variation (Mendenhall et al., 1993). A good analysis of
weights includes making histograms of the data to check the form of their
distribution. The measures of a distributions deviation could be used, like standard
deviation, coefcient of variation, skewness, and kurtosis. The small values of these
measures (close to 0) indicate the consensus among the evaluators. Large negative
kurtosis shows that the evaluators are far from consensus. If the data are not
normally distributed, and the measures of spread are not small, the average value is
not representative. If the measures of shape are large (negative kurtosis) for the
criteria weights, sensitivity analysis covering the range of weights should be
performed within MCDM procedure.
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 176
4. Multicriteria ranking and compromise solution
Among the numerous approaches available for conict management, MCDM is
one of the most widely used. In this approach, practical problems are often
characterized by several non-commensurable and competing (conicting) criteria,
and there is no solution satisfying all the criteria simultaneously. Thus, a
compromise solution for problems with conicting criteria should be determined
to help decision makers reach a nal decision.
The MCDM procedure applied in this paper consists of the following steps:
(a) Establishing system evaluation criteria (multiple) that relate system capabilities
to goals;
(b) Developing (designing) alternative systems for attaining the goals (generating
alternatives);
(c) Evaluating alternatives in terms of criteria (the values of the criterion functions);
(d) Applying a normative multicriteria analysis method (such as compromise
ranking);
(e) Accepting one alternative as optimal (preferred); and
(g) If the nal solution is not accepted, then gather new information and go into the
next iteration of multicriteria optimization.
Steps (a) and (e) are performed at the decision (upper) level, where decision
makers have a central role. Other steps are mostly mechanical tasks.
Alternatives can be generated and their feasibility can be tested using
mathematical models, physical models, and/or by experiments on the existing
system or other similar systems. Generating alternatives may be a very complex
process, there is no general procedure or model, and no mathematical procedure
could replace the human creativity in generating and evaluating alternatives.
Constraints are seen as high-priority objectives, and they must be considered and
satised in the alternatives generating process.
Assuming that each alternative is evaluated according to each criterion f
unction, the compromise ranking method could be applied to determine a
compromise solution, helping the decision makers to reach nal decision.
Compromise solution is a feasible solution which is the closest to the ideal. A
compromise means an agreement established by mutual concessions. The
compromise ranking method (known as VIKOR) has been introduced as one
applicable technique to implement within MCDM. The main features of this method
are presented in Appendix B.
By the compromise ranking method a compromise solution is determined which
could be accepted by the decision makers because it provides a maximum group
utility of the majority (with measure S; representing concordance), and a
minimum of individual regret for the opponent (with measure R; representing
discordance). The VIKOR algorithm determines the weight stability intervals for
the obtained compromise solution with the input weights given by experts
(Opricovic, 1998).
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 177
5. An illustrative example
The Takarazima Japanese Siki Kisegi cuisine restaurant was established in 1988.
The restaurant is about 600 m
2
and has about 60 employees. It is known for its ne
cuisine and high-quality service. The restaurant managers decided to open another
restaurant, considering four alternative restaurant locations in Taipei, spatially
located as presented in Fig. 1.
Alternative A is in the Tien Mou commercial center on a side street of Sec. 6 of
Chung Chang North Road. The site is near Ta Yeh, a very big department store.
According to the study of Compiled Investigation Information of Taipei
Commercial Centers (Money Culture Enterprise Company, Limited, 1995), the
Tien Mou commercial center is a compound commercial type. In addition, many of
the foreign people in Taipei live there, and many embassies are also located there.
Alternative B is in the Chung Hsiao commercial center on a side street of Sec. 4 of
Chung Hsiao East Road (Fig. 2). The site is near SOGO department store, the
largest in Taipei, and Tong Lin another big department store. The Chung Hsiao
commercial center is a recently developed commercial center, still expanding. The
commercial center is located at the intersection of two subway lines, and it is a
metropolitan core type of commercial center. This commercial area is B1.5 km long
extending from east to west, and 0.75 km wide from south to north.
Alternative C is located in the Chung Shan commercial center on a side street of
Sec. 3 of Chung Shan North Road, and near the Chin Kwong market. The Chung
Shan commercial center is formed within the enclosure of Chin Kwong market to the
peripheral areas. Since most of the foreigners in Taipei live in Tien Mou, Sec. 3 of
Chung Shan North Road, which leads there, it was developed as a special area.
Many stores located in this area sell imported goods and attract both foreign and
local people for shopping.
Fig. 1. Restaurant locations in Taipei.
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 178
Alternative D is located in the Hsin Yi commercial center on Sec. 4 of Hsin Yi
Road (Fig. 3). The site is near the Taipei World Trade Center and the Taipei City
Government Hall. The Hsin Yi commercial center forms the core of the Hsin Yi
Planning District, which is a new city political center and is famous for the high
quality of its public facilities and housing. The Hsin Yi commercial center is an
international metropolitan type of commercial center.
The criteria can be divided into two groups: quantitative and qualitative.
Quantitative criteria are rent cost, transportation cost, convenience of mass
Fig. 2. Alternative B:
Fig. 3. Alternative D:
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 179
transportation system, size of parking space, pedestrian volume, number of
competitors and the size of commercial area where the restaurant will be located.
The performance values of quantitative criteria were determined through eld
investigation. Qualitative criteria are those that cannot be quantied or involve
subjective estimation, including the intensity of competitors, convenience of garbage
disposal and sewage capacity. The performance values of qualitative criteria are
determined by the decision-making group of the Takarazima restaurant. The
performance values for each of the criteria are presented in Table 4.
The compromise ranking method presented in this paper was applied with criteria
values from Table 4, and weights from Table 3. The obtained ranking list (by
VIKOR) is presented in Table 5 (with weights total). The same ranking is obtained
with all four sets of weights from Table 3.
Using the VIKOR computer program, the weight stability intervals for the
obtained compromise solution are obtained, as follows:
0:000pw
1
p1:000; 0:000pw
2
p0:571;
0:000pw
3
p0:430; 0:000pw
4
p0:125;
0:000pw
5
p0:183; 0:000pw
6
p0:148;
0:000pw
7
p0:464; 0:000pw
8
p1:000;
0:000pw
9
p1:000; 0:0000pw
10
p0:113;
0:000pw
11
p0:112:
The weight stability intervals are determined for each criterion separately. The
obtained intervals show that the obtained compromise solution is the same for all
four sets of weights from Table 3, and all values of weights are within the
corresponding interval.
The result by method VIKOR is the set of compromise solutions including
alternative B (Fig. 2) and alternative D (Fig. 3), showing that there is no acceptable
Table 4
Criteria values for alternatives
Criteria Alternatives
A B C D
Rent cost (NT dollars/m
2
per month) 4000 3500 6000 4500
Transportation cost (NT dollars per month) 12,000 8000 7000 8500
Convenience of mass transportation system 10 20 26 17
Parking capacity (park. units) 3650 347 742 1354
Pedestrian volume (people per day) 5508 10,439 17,803 15,840
Number of competitors 9 16 18 8
Intensity of competition 0.4 0.467 0.723 0.667
Size of commercial area (km
2
) 0.75 1.06 0.81 0.84
Extent public facilities 0.5 0.667 0.533 0.633
Convenience of garbage disposal 0.7 0.4 0.467 0.667
Sewage capacity 0.633 0.3 0.433 0.6
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 180
advantage of B: At this point, alternative B may be proposed to the decision maker,
with the remark that alternatives B and D are both possible (they are close).
The results of comparing alternatives B and D are presented in Table 6, indicating
that alternative B is better than D f
B
gf
D
according to the two most important
criteria (rent cost and size of commercial area).
6. Conclusions
This multicriteria optimization method may be applied to help decision makers
achieve acceptable compromises in many situations, although the established set of
criteria is related to the Taipei case. By the compromise ranking method, the
compromise solution is determined which would be most acceptable to the decision
makers because it provides a maximum group utility for the majority, and a
minimum of individual regret for the opponents.
The evaluation of objective importance indicates that the transportation aspect is
more important in this problem of restaurant location. The weight value for rent cost
at the attribute level is the greatest, showing that all the decision makers have
estimated it with the highest importance. The consensus for the evaluation criteria is
Table 6
A comparison of alternatives B and D
Criteria f
B
f
D
f
B
gf
D
w
Rent cost (NT dollars/m
2
per month) 3500 4500 +1000 0.1808
Transportation cost (NT dollars per month) 8000 8500 +500 0.0502
Convenience of mass transportation system 20 17 +3 0.0739
Parking capacity (park. units) 347 1354 1007 0.0912
Pedestrian volume (people per day) 10,439 15,840 5401 0.0945
Number of competitors 16 8 8 0.1026
Intensity of competition 0.467 0.667 +0.2 0.0581
Size of commercial area (km
2
) 1.06 0.84 +0.22 0.1779
Extent of public facilities 0.667 0.633 +0.034 0.0756
Convenience of garbage disposal 0.4 0.667 0.267 0.0492
Sewage capacity 0.3 0.6 0.3 0.0457
Table 5
Results of ranking alternatives
Rank Alternative (Q-measure)
1 B 0.000
2 D 0.247
3 A 0.711
4 C 1.000
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 181
the highest for the commercial area where the restaurant is located, indicating that
all evaluators are more consistent with this evaluation criterion.
The result of multicriteria ranking of alternative restaurant locations is the set of
compromise solutions including alternatives B and D (they are close, ranking:
BEDgAgC). It may be recommended to the decision maker that alternatives B
and D are both good locations. The alternative B is in an expanding commercial
center at an intersection of two subway lines, and it is a metropolitan core type of
commercial center. The alternative D is in Hsin Yi commercial center and near the
Taipei World Trade Center and the Taipei City Government Hall, a new city
political center in this area.
Acknowledgements
This paper is partly a result of the project NSC88-2811-H009-0001, supported by
the National Science Council of Taiwan. The paper has been prepared with the
generous assistance of the general manager of Takarazima Japanese Siki Kisegi
cuisine restaurant, Mr. Shen Chin-Chung, and we are especially grateful for his
assistance. The constructive comments of the editor and the referees are gratefully
acknowledged.
Appendix A
The AHP is a systematic procedure for representing the elements of a problem,
hierarchically. The AHP method was developed by Saaty in 1971 (Saaty, 1980; Saaty
and Vargas, 1982). The procedures used in this paper may be summarized as follows:
(1) Structuring the hierarchy for evaluation. The AHP method is used to make the
decomposition (or structuring) of the problem as a hierarchy. In general, the AHP
method divides the problem into three levels: (a) the goal for resolving the problem;
(b) the objectives for achieving the goal; and (c) the evaluation criteria for each
objective. An example is presented in Table 1.
(2) Constructing the pairwise comparison matrix. After structuring a hierarchy, the
pairwise comparison matrix for each level is constructed. During the pairwise
comparison, the nominal scale is used for evaluation.
The scale of relative importance is presented in Table 7. In the evaluation process
(by experts), the AHP questionnaire sheet for restaurant location is used. The sheet
for the objective level is presented in Table 8, it is used in comparing objectives, one
on the left with one on the right.
(3) Calculating the weights and testing the consistency for each level. For each
pairwise comparison matrix A; using the theory of eigenvector, i.e.
(A l
max
I)w 0; to calculate the eigenvalue l
max
and the eigenvector
w w
1
; w
2
; :::; w
n
; weights can be estimated. Finally, the consistency of the
comparison matrix was tested and the opinions of the regional decision-maker
group were integrated. In the consistency test, consistency index (CI) is utilized to
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 182
determine the degree of consistency, generally speaking, when CIo0.1 it is
considered to be acceptable.
Appendix B
The compromise ranking method (known as VIKOR) has been introduced as one
applicable technique to implement within MCDM (Opricovic, 1998). Assuming that
each alternative is evaluated according to each criterion function, the compromise
ranking could be performed by comparing the measure of closeness to the ideal
alternative. The multicriteria merit for compromise ranking is developed from the
L
p
-metric used in compromise programming method (Yu, 1973; Zeleny, 1982). The
various alternatives are denoted as a
1
; a
2
; :::; a
J
: For an alternative a
j
the merit of the
ith aspect is denoted by f
ij
; i.e. f
ij
is the value of the ith criterion function for the
alternative a
j
; and n is the number of criteria.
Fig. 4 illustrates ideal F

1
; f

2
and compromise solution F
c
f
c
1
; f
c
2
within
a bicriteria problem; a compromise as an agreement established by mutual
concessions is illustrated by f
c
1
pf

1
and f
c
2
pf

2
:
Table 7
Scale of relative importance
Intensity of
importance
Denition Explanation
1 Equal importance Two activities contribute equally to the
objective
2 Intermediate between equal and weak
3 Weak importance of one over another Experience and judgment slightly favor
one activity over another
4 Intermediate between weak and strong
5 Essential or strong importance Experience and judgment strongly favor
one activity over another
6 Intermediate between strong and
demonstrated
7 Demonstrated importance An activity is strongly favored and its
dominance is demonstrated in practice
8 Intermediate between demonstrated and
absolute
9 Absolute or extreme importance The evidence favoring one activity over
another is of the highest possible order
of afrmation
Reciprocals of
above non-
zero numbers
If activity i has one of the above non-
zero numbers assigned to it when
compared with activity j; then j has the
reciprocal value when compared with i:
Reasonable assumption
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 183
Compromise programming method introduced L
p
-metric as an aggregated
function. The development of VIKOR method started with the following form of
L
p
-metric:
L
p; j

X
n
i1
w
i
f

i
f
ij
= f

i
f

i

p
( )
1=p
; 1pppN; j 1; 2; :::; J:
Within VIKOR method L
1; j
(as S
j
) and L
N;j
(as R
j
) are used in formulating ranking
merit (boundary solutions). The solution obtained by min
j
S
j
is with a maximum
Table 8
AHP questionnaire sheet for restaurant location
Level 1: Absolute Strong Equal Strong Absolute Level 1:
Objective importance importance importance importance importance Objective
Economics

9:1 7:1 5:1 3:1 1:1 1:3 1:5 1:7 1:9 Transportation
8:1 6:1 4:1 2:1 1:2 1:4 1:6 1:8
Economics 9:1 7:1 5:1 3:1 1:1 1:3 1:5 1:7 1:9 Competition
8:1 6:1 4:1 2:1 1:2 1:4 1:6 1:8
Economics 9:1 7:1 5:1 3:1 1:1 1:3 1:5 1:7 1:9 Commercial
8:1 6:1 4:1 2:1 1:2 1:4 1:6 1:8 area
Economics 9:1 7:1 5:1 3:1 1:1 1:3 1:5 1:7 1:9 Environment
8:1 6:1 4:1 2:1 1:2 1:4 1:6 1:8
Transportation 9:1 7:1 5:1 3:1 1:1 1:3 1:5 1:7 1:9 Competition
8:1 6:1 4:1 2:1 1:2 1:4 1:6 1:8
Transportation 9:1 7:1 5:1 3:1 1:1 1:3 1:5 1:7 1:9 Commercial
8:1 6:1 4:1 2:1 1:2 1:4 1:6 1:8 area
Transportation 9:1 7:1 5:1 3:1 1:1 1:3 1:5 1:7 1:9 Environment
8:1 6:1 4:1 2:1 1:2 1:4 1:6 1:8
Competition 9:1 7:1 5:1 3:1 1:1 1:3 1:5 1:7 1:9 Commercial
8:1 6:1 4:1 2:1 1:2 1:4 1:6 1:8 area
Competition 9:1 7:1 5:1 3:1 1:1 1:3 1:5 1:7 1:9 Environment
8:1 6:1 4:1 2:1 1:2 1:4 1:6 1:8
Commercial 9:1 7:1 5:1 3:1 1:1 1:3 1:5 1:7 1:9 Environment
area 8:1 6:1 4:1 2:1 1:2 1:4 1:6 1:8
Fig. 4. Illustrating compromise.
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 184
group utility (majority rule), and the solution obtained by min
j
R
j
is with a
minimum individual regret of the opponent.
Weighting coefcients (weights w
i
) are introduced to express the relative
importance of the criteria. The weights have no clear economic meaning, but the
use of weights gives the opportunity for modeling the real decision making.
The compromise ranking algorithm VIKOR has the following steps:
(a) Determination of the best f

i
and the worst f

i
values of all criterion functions,
i 1; 2; :::; n: If the ith function represents a benet, then
f

i
max
j
f
ij
; f

i
min
j
f
ij
:
(b) Compute the values S
j
and R
j
; j 1; 2; :::; J; by the relations
S
j

X
n
i1
w
i
f

i
f
ij
= f

i
f

i
;
R
j
max
i
w
i
f

i
f
ij
= f

i
f

i
;
where w
i
are the weights of criteria.
(c) Compute the values Q
j
; j 1; 2; :::; J; by the relation
Q
j
vS
j
S

=S

1 vR
j
R

=R

;
where
S

min
j
S
j
; S

max
j
S
j
;
R

min
j
R
j
; R

max
j
R
j
:
v is introduced as weight of the strategy of the majority of criteria (or the
maximum group utility), usually v=0.5.
(d) Rank the alternatives, sorting by the values S; R; and Q: The results are three
ranking lists.
(e) Propose as a compromise solution the alternative a
0
which is the best ranked
one by the measure Q if the following two conditions are satised:
C1. Acceptable Advantage: Qa
00
Qa
0
XDQ where a
00
is the alternative with
second position in the ranking list by Q; DQ 1=J 1; J is the number of
alternatives. (DQ 0:25 if Jp4:)
C2. Acceptable stability in decision making: The alternative a
0
must also be best
ranked by S or/and R: This compromise solution is stable within a decision-making
process, which could be voting by majority rule (when v > 0:5 is needed), or by
consensus vE0:5; or with veto vo0:5: Here, v is the weight of the decision-
making strategy the majority of criteria (or the maximum group utility).
If one of the conditions is not satised, then a set of compromise solutions is
proposed, which consists of
*
Alternatives a
0
and a
00
if only the condition C2 is not satised, or
*
Alternatives a
0
; a
00
; y; a
k
if the condition C1 is not satised; a
k
is determined by
the relation Qa
k
Qa
0
EDQ (the positions of these alternatives are in
closeness).
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 185
The best alternative, ranking by Q; is one with the minimum value of Q:
The main ranking result is the compromise ranking list of alternatives, and the
compromise solution with the advantage rates.
Ranking by this algorithm may be performed with different values of criteria
weights w
i
; analyzing the impact of criteria weights on proposed compromise
solution. The VIKOR algorithm determines the weight stability intervals, for the
obtained compromise solution with the input weights, indicating the preference
stability of obtained compromise solution. This is a helpful tool in MCDM,
particularly in the situation when the decision maker is not able to express preference
at the beginning of system design.
References
Applebaum, W., Cohen, S.B., 1960. Evaluating store sites and determining store rents. Economic
Geography 36 (1), 135.
Money Culture Enterprise Company, Limited, 1995. Compiled Investigation Information of Taipei
Commercial Centers.
Guerrier, Y., Deery, M., 1998. Research in hospitality human resource management and organizational
behavior. International Journal of Hospitality Management 17 (2), 145160.
ITT Project, 1998. Effect of Bus Lane on Level of Service of Bus, Bus Ridership and Trafc Congestion
Taking Taipei as an Example, NSC87-2211-E009-011. Institute of Trafc and Transportation, NCTU,
Taipei.
Jain, A.K., Mahajan, V., 1979. Evaluating the Competitive Environment in Retailing Using Multiplicative
Competitive Interactive Models, Research in Marketing. JAI Press, Greenwich, CT.
Kem, J., Simmons, J., 1990. The Retail Environment. Routledge, London.
Mendenhall, W., Reinmuth, J., Beaver, R., 1993. Statistics for Management and Economics. Wadsworth,
Belmont, CA.
Nelson, R.L., 1958. The Selection of Retail Location. McGraw-Hill, New York.
Olsen, M., West, J., Tse, E.C.Y. (Eds.), 1998. Strategic Management in the Hospitality Industry. Wiley,
New York.
Opricovic, S., 1998. Multicriteria Optimization in Civil Engineering. Faculty of Civil Engineering,
Belgrade.
Saaty, T.L., 1980. The Analytic Hierarchy Process. McGraw-Hill, New York.
Saaty, T.L., Vargas, L.G., 1982. The Logic of Priorities: Application in Business, Energy, Health, and
Transportation. Kluwer-Nijhoff, Boston.
THI Study, 1996. The Study on the Analysis of Parking Demand and Supply Model. THI Consult Co.,
Taipei.
Yu, P.L., 1973. A class of solutions for group decision problems. Management Science 19 (8), 936946.
Yun, L.L., Hing, N., 1995. Measuring quality in restaurant operations: an application of the SERVQUAL
instrument. International Journal of Hospitality Management 14 (34), 293310.
Zeleny, M., 1982. Multiple Criteria Decision Making. McGraw-Hill, New York.
Gwo-Hshiung Tzeng is a National Distinguished Chair-Professor in the Department
of Management of Technology at the National Chiao Tung University, Hsinchu,
Taiwan.
Mei-Hwa Teng is a lecturer at the National Open University, Taipei. She was a
graduate student at the National Chiao Tung University.
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 186
June-Jye Chen is an Assistant Professor in the Department of Marketing and
Distribution Management at the Ling Tung College, Taichung. He was an Assistant
Professor in the Department of Hotel Management, Ming-Shin Institute of
Technology, 1 Hsin-Hsing Road, Hsin-Fong, HsinChu (when this paper was done).
Seram Opricovic is a Professor at the Faculty of Civil Engineering, University of
Belgrade, Yugoslavia, Visiting Professor at National Chiao Tung University, and is
supported by National Science Council of Taiwan, in 1999/2000 and 2001/2002.
G.-H. Tzeng et al. / Hospitality Management 21 (2002) 171187 187