17
11.04.201 11.04.201 3 1 Dr. BOTTI Laurent Dr. BOTTI Laurent Dr. PEYPOCH Nicolas Dr. PEYPOCH Nicolas Perpignan University / CAEPEM Perpignan University / CAEPEM 77th meeting of the European Group on MCDA 77th meeting of the European Group on MCDA 11th April 2013, University of Rouen 11th April 2013, University of Rouen

Botti peypoch mcda

Embed Size (px)

DESCRIPTION

Conférence MCDA IAE de Rouen Avril 2013

Citation preview

Page 1: Botti peypoch mcda

11.04.201311.04.2013

11

Dr. BOTTI Laurent Dr. BOTTI Laurent Dr. PEYPOCH NicolasDr. PEYPOCH Nicolas

Perpignan University / CAEPEMPerpignan University / CAEPEM

77th meeting of the European Group on MCDA 77th meeting of the European Group on MCDA

11th April 2013, University of Rouen 11th April 2013, University of Rouen

Page 2: Botti peypoch mcda

11.04.201311.04.2013

22

1.1. IntroductionIntroduction

2.2. MCDA methods and destination competitivenessMCDA methods and destination competitiveness

3.3. ELECTRE I and TDC : the case of Hawaiian islandsELECTRE I and TDC : the case of Hawaiian islands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

Multi-Criteria ELECTRE Multi-Criteria ELECTRE

Method and Destination Method and Destination

Competitiveness Competitiveness

Page 3: Botti peypoch mcda

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

11.04.201311.04.2013

33

The global tourism industry is The global tourism industry is dynamicdynamic

CompetitivenessCompetitiveness is a growing interest area is a growing interest area

Page 4: Botti peypoch mcda

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

11.04.201311.04.2013

44

Due to their centrality in the tourism system, Due to their centrality in the tourism system, tourism tourism

destinationsdestinations (TD) benefit from this interest (Cracolici & (TD) benefit from this interest (Cracolici &

Nijkam, 2008) Nijkam, 2008)

To understand TD competitiveness, we dispose of the To understand TD competitiveness, we dispose of the

Ritchie & Crouch (2003) modelRitchie & Crouch (2003) model, the most cited one, the most cited one

This model integrates all the relevant This model integrates all the relevant factorsfactors that might that might

typify the competitiveness of a destinationtypify the competitiveness of a destination

This paper aims to show the interest of MCDA methods This paper aims to show the interest of MCDA methods

regarding the regarding the operationalizationoperationalization of this model of this modelLet’s move to section 2

Page 5: Botti peypoch mcda

11.04.201311.04.2013

55

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

Why are MCDA methods relevant to deal with competitiveness of Why are MCDA methods relevant to deal with competitiveness of

tourism destination ? tourism destination ?

MCDA for methods providing quantitative approach to support MCDA for methods providing quantitative approach to support

decision making in problems involving decision making in problems involving several criteria and several criteria and

choiceschoices (alternatives or actions) (Figueira, Mousseau & Roy, (alternatives or actions) (Figueira, Mousseau & Roy,

2005) 2005)

TD are an integrated TD are an integrated set of tourist facilitiesset of tourist facilities which have to face which have to face

the challenge of operating their resources effectively and the challenge of operating their resources effectively and

efficiently in order to supply an efficiently in order to supply an experienceexperience that outperforms that outperforms

alternativealternative experiences on the market (Ritchie & Crouch, 2000) experiences on the market (Ritchie & Crouch, 2000)

Page 6: Botti peypoch mcda

11.04.201311.04.2013

66

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

Tourists who whish to enjoy a satisfying experience try to Tourists who whish to enjoy a satisfying experience try to

selectselect one destination from a set of one destination from a set of nn possible alternatives possible alternatives

and on the basis of and on the basis of mm criteria criteria

The main framework to understand The main framework to understand TDCTDC is that of Ritchie is that of Ritchie

& Crouch (2003) & Crouch (2003)

– based on five competitiveness components (based on five competitiveness components (criteriacriteria))

Attractors (Climate, History…) Attractors (Climate, History…)

Supporting factors (Accessibility, Hospitality…) Supporting factors (Accessibility, Hospitality…)

Destination management (Marketing, HRM…) Destination management (Marketing, HRM…)

Destination planning (Positioning, Branding…) Destination planning (Positioning, Branding…)

Amplifying determinants (Safety…) Amplifying determinants (Safety…)

Page 7: Botti peypoch mcda

11.04.201311.04.2013

77

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

=> => Destination selectionDestination selection can be seen as a MCDA problem can be seen as a MCDA problem

Let’s move to an

application

Page 8: Botti peypoch mcda

11.04.201311.04.2013

88

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

Although MCDA methods can be applied to different Although MCDA methods can be applied to different

areas, the litterature is quite areas, the litterature is quite narrownarrow when when

considering the tourism field considering the tourism field

– TOPSISTOPSIS was used by Zhang et al. (2011) to rank 16 was used by Zhang et al. (2011) to rank 16

cities in Chinacities in China

– TOPSIS, TOPSIS, PROMETHEEPROMETHEE and the and the WSMWSM was used by was used by

Ishizaka, Nemery and Lidouh (2013) to select the Ishizaka, Nemery and Lidouh (2013) to select the

location of a casino in Londonlocation of a casino in London

– ELECTRE IIELECTRE II was used by Andrades-Caldito et al. (2013) was used by Andrades-Caldito et al. (2013)

to rank provinces of Andalusia (Spain)to rank provinces of Andalusia (Spain)

Page 9: Botti peypoch mcda

11.04.201311.04.2013

99

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

Here, Here, ELECTRE IELECTRE I (Roy, 1991) is applied to (Roy, 1991) is applied to choosechoose the the

best destination from a given set of alternativesbest destination from a given set of alternatives

Widely used, the MCDA Widely used, the MCDA outrankingoutranking method ELECTRE I method ELECTRE I

should be applied when should be applied when all criteriaall criteria are coded in are coded in

numerical scalesnumerical scales with with identical ranges identical ranges (Figueira, (Figueira,

Mousseau, Roy, 2005)Mousseau, Roy, 2005)

– Data are derived from the 2011 Hawaiian VSA Data are derived from the 2011 Hawaiian VSA ReportReport

which presents results of a which presents results of a surveysurvey conducted by the HTA conducted by the HTA

– The HTA report attempts to portray visitor’s evaluation The HTA report attempts to portray visitor’s evaluation

of their of their experienceexperience by covering by covering various aspectsvarious aspects of their of their

trip for the island that they stayed the longesttrip for the island that they stayed the longest

Page 10: Botti peypoch mcda

11.04.201311.04.2013

1010

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

Visitors were asked to Visitors were asked to rate few attributes of islandsrate few attributes of islands : :

accomodations, restaurant, shopping, golf, attractions, accomodations, restaurant, shopping, golf, attractions,

transportation, airports, parks & beaches transportation, airports, parks & beaches

– We use the We use the proportion of very satisfiedproportion of very satisfied visitors as visitors as

measurement of performancemeasurement of performance

– Our application focuses Our application focuses

only on visitors from only on visitors from

EuropeEurope in 2011 in 2011

– 6 major islands but due 6 major islands but due

to data availability, we to data availability, we

focus on focus on 4 islands4 islands (A = (A =

aaii ; i = 1,…,4) ; i = 1,…,4)

Page 11: Botti peypoch mcda

11.04.201311.04.2013

1111

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

44 criteria, their weights (w criteria, their weights (w jj, from , from Crouch, 2011Crouch, 2011) and ) and attributes attributes : :

Considering FConsidering Fjj = {f = {ftjtj ; t=1,…,p}, the set of ; t=1,…,p}, the set of p p attributes of attributes of

criterion criterion jj, performances are obtained by : , performances are obtained by :

Page 12: Botti peypoch mcda

11.04.201311.04.2013

1212

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

PerformancePerformance of alternatives on each criterion : of alternatives on each criterion :

Page 13: Botti peypoch mcda

11.04.201311.04.2013

1313

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

Concordance and discordance matrixes were Concordance and discordance matrixes were

performed with performed with LINAM softwareLINAM software (Logiciel Interactif (Logiciel Interactif

d’Analyse Multicritère by P. Wieser from Lausanne) d’Analyse Multicritère by P. Wieser from Lausanne)

Second step of ELECTRE I is the derivation of a Second step of ELECTRE I is the derivation of a

recommendation based on the outranking relations i.e. recommendation based on the outranking relations i.e.

identify identify a small as possible subset of actionsa small as possible subset of actions, from , from

which the best compromise action(s) could be which the best compromise action(s) could be selectedselected

Construction of a graph considering the concordance Construction of a graph considering the concordance

level level C*C* and the discordance level and the discordance level D*D*

Page 14: Botti peypoch mcda

11.04.201311.04.2013

1414

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

Let’s move to the last section

For C* = 0.56 and D*= For C* = 0.56 and D*=

0.42 0.42

For C* = 0.56 and D*= For C* = 0.56 and D*=

0.26 0.26

=>=> O’ahu must be choosen by tourists

This island is the most competitive

By considering that the satisfaction achieved by tourists can be used to indirectly evaluate the competitive ability of TD to outperform others

destinations

Page 15: Botti peypoch mcda

11.04.201311.04.2013

1515

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

Interests : Interests :

– We show how MCDA outranking approach ELECTRE I can We show how MCDA outranking approach ELECTRE I can

be used to analyse TDC be used to analyse TDC =>=>FirstFirst application of ELECTRE application of ELECTRE

I to this field I to this field

– QuantitativeQuantitative operationalization of the C&R model operationalization of the C&R model

Limitations : Limitations :

– Performances are obtained via an Performances are obtained via an arithmeticarithmetic mean mean

– Data can be considered as Data can be considered as reductivereductive as we focus on as we focus on

European visitorsEuropean visitors

– What about What about indifferenceindifference ? What about preference ( ? What about preference (strongstrong

and and weakweak) ? ) ?

Page 16: Botti peypoch mcda

11.04.201311.04.2013

1616

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

Perspectives : Perspectives :

– RankRank all the destination belonging to a given set of all the destination belonging to a given set of

alternatives from the best to the worst (ski resorts, alternatives from the best to the worst (ski resorts,

seaside resorts, cities…) with ELECTRE II or ELECTRE III seaside resorts, cities…) with ELECTRE II or ELECTRE III

– CompareCompare ELECTRE ELECTRE rankingsrankings with other rankings – for with other rankings – for

example efficiency ranking (obtained with DEA method example efficiency ranking (obtained with DEA method

or others) or others)

Page 17: Botti peypoch mcda

11.04.201311.04.2013

1717

1.1. IntroductionIntroduction

2.2. MCDA methods and TDCMCDA methods and TDC

3.3. ELECTRE I and TDC : the case of Hawaiian ELECTRE I and TDC : the case of Hawaiian islandsislands

4.4. Interests, limitations and perspectivesInterests, limitations and perspectives

Thank you for attention!Thank you for attention!

[email protected]@univ-perp.fr

[email protected]@univ-perp.fr