1 Introduction
2 AHP/ANP for Supporting Strategic Decisions
3 Literature-Based Overview of Group Aggregation Techniques for AHP/ANP
Category I | |
---|---|
Aggregation by means of direct information | |
Aggregation of individual judgments (AIJ)
| |
Representation (and evaluation) [9]; [18]; [20]; [33]; [38] | |
Case study: ANP and AIJ (WGMM) [16]; [28] | |
Analyses of consistency in AIJ-approaches [4]; [23]; [38]; [45]; [48] | |
Hybrid approach that takes advantages of AIJ and AIP (with weighted least square measures) [37] | |
Aggregation of individual priorities (AIP)
| |
Representation (and evaluation) [14]; [18]; [20]; [33]; [38] | |
AIP (WAMM) for multiple-issues [14] | |
Case study: consens building with AIP and discussion [26] | |
Case study: AHP and AIP (WAMM) [32] | |
Multiplicative AHP and AIP [36]; [41] | |
Hybrid approach that takes advantages of AIJ and AIP (with weighted least square measures) [37] |
Category II | |
---|---|
Aggregation by means and indirect information | |
A loss function approach to group preference aggregation [13] | |
Visualizing group decisions using Sammon-Maps [15] | |
Aggregation of individual preference structures (AIPS) [19] | |
Group AHP with aggregation on preferential differences and rankings (Group AHP model) [25] |
Category III | |
---|---|
Deriving priorities and/or aggregating priorities using mathematical extension | |
Fuzzy AHP/ANP
| |
Group ANP approach using quadratic mathematical programming and interval preference information [27] | |
AHP group prioritization by fuzzy preference programming method [29] | |
AHP and fuzzy preference relations [42] | |
Fuzzy ANP and nonlinear programming [46] | |
Fuzzy AHP and goal programming [47] | |
Group preference building by using a Bayesian approach
| |
Bayesian priorization procedure for AHP-group decision [2] | |
A Bayesian approach for consensus building in (multiplicative) AHP-group decision making [3] | |
A Bayesian approach based on mixtures for multiplicative group-AHP [21] | |
Aggregation by using alternativ mathematical functions
| |
Deterministic and stochastic approach for group aggregation, depending on the group size (maximum likelihood method) [7] | |
AHP group decision making to reduce investment risks using method of least squares [44] | |
(Other) alternativ methods for deriving priorities
| |
Statistical method for deriving priorities that allows dependence among the entries of the pairwise comparison matrix [6] | |
Consensus models for AHP group decision making under row geometric mean priorization method [17] | |
Consistency consensus matrix (CCM)-approach based on preference structures distribution and consistency stability intervals [30]; [31] |
Category IV | |
---|---|
Extensions of AHP/ANP | |
Group preference building by combination with additional OR-methods
| |
AHP and goal programming with priority interval vectors [10] | |
Weighted geometric mean data envelopment analysis (DEA)-model [22] | |
Linear programming method to generate weights in AHP [24] | |
Group fuzzy preference programming method (GFPP) [29] | |
DEA model for group decision making in AHP [43] | |
Fuzzy AHP and goal programming [47] |
Category V | |
---|---|
Specific variants of AHP | |
Multiplicative AHP
| |
A Bayesian approach for multiplicative group-AHP [3]; [21] | |
Group aggregation in the multiplicative AHP with SMART (additive weighting method) under consideration of power relations [5]; [40] | |
Linking (multiplicative) AHP with social choice method (AIP-approach) [36] | |
Stochastic group preference model for multiplicative AHP with relative decisional power values [39] | |
AIJ and AIP-approaches in the multiplicative AHP [41] |
Category VI | |
---|---|
General aspects on multipersonal AHP/ANP | |
Rational, formal and/or mathematical analyses
| |
Procedures and requirements for synthesizing ratio judgments [1] | |
Analyses of group welfare functions/ dispersion of group judgments [34]; [135] | |
Analyses of consistency
| |
Experiment on the consistency of AIJ-matrices (Monte-Carlo-simulations) [4] | |
Analysis of the stability of consistency in AIJ-approach [23]; [38]; [45]; [48] | |
Methods for deriving group members’ weights
| |
Method for group AHP with the non-equivalent importance of individuals in the group [8] | |
Intrinsic process for deriving members’ weightages (by using a specific form of the eigenvector-approach) [33] | |
Allocation of group members’ weights in the multiplicative AHP and SMART [40] | |
Other analyses
| |
The effect of group-AHP on group polarization with focus on “social comparison (SC)” and “persuasive arguments (PA)” [11] | |
Evaluation and possible adjustments of the AHP to various contexts of GDM [18] | |
A cost constrained mediation model for AHP negotiated decision making [12] |
4 Evaluation
4.1 Case Example Description
\(\textit{DM}_{r}\)
|
\(\textit{DM}_{1}\)
|
\(\textit{DM}_{2}\)
|
\(\textit{DM}_{3}\)
|
---|---|---|---|
\(A_i\)
|
\(A_{1}\quad \quad \,A_{2}\quad \quad \,\,A_{3}\quad \quad \,\,A_{4}\)
|
\(A_{1}\quad \quad \,A_{2}\quad \quad \,\,A_{3}\quad \quad \,\,A_{4}\)
|
\(A_{1}\quad \quad \,A_{2}\quad \quad \,\,A_{3}\quad \quad \,\,A_{4}\) |
\(P_i^{gl}(\textit{DM}_{r})\)
| 0.236 0.418 0.164 0.181 | 0.490 0.127 0.173 0.210 | 0.238 0.262 0.063 0.437 |
Ranking |
\(A_{2}\, > \,\,A_{1}\, > \,\,A_{4}\, > \,\,A_{3}\)
|
\(A_{1}\, >\,\,A_{4}\, >\,\, A_{3}\, >\,\,A_{2}\)
|
\(A_{4}\, >\,\,A_{2}\, >\,\,A_{1}\, >\,\,A_{3}\)
|
4.2 Evaluation Criteria
4.3 Representation and Evaluation of Appropriate Group Aggregation Techniques
4.3.1 Representation
Step 1: Computation of \(\overline{\textit{CR}} \) and \(\textit{VAR}_{CR} \)
|
\(\overline{CR} =\mathop \sum \nolimits _{r=1}^R \frac{\textit{CR}_r }{R}; \textit{VAR}_{CR} =\mathop \sum \nolimits _{r=1}^R \frac{\left( {\textit{CR}_r -\overline{\textit{CR}} } \right) ^{2}}{\left( {R-1} \right) },\) where \(\overline{\hbox {CR}} \) is the mean of the inconsistency ratio and \(\textit{VAR}_{\textit{CR}} \) is the variance of the inconsistency ratio |
Step 2: Calculation of the expected loss X
|
\(X=\textit{VAR}_{\textit{CR}} +\left( {\overline{\textit{CR}} } \right) ^{2}\)
|
Step 3: Determination of a collective weight of the group by using the evaluation reliability function \(F\left( X \right) \)
|
\(F(\hbox {X})=\left\{ \begin{array}{l} {1,} \\ {\exp \left( {-{\upomega }\,\hbox {X}} \right) ,} \\ {0,} \\ \end{array}\right. \)
\(\begin{array}{l} {X=0,} \\ {0<X<\textit{tolerance}\,\textit{limit}} \\ {X\ge \textit{tolerance}\,\textit{limit},} \\ \end{array},\) where \({\upomega }\) is a coefficient for each matrix dimension. Tolerance limit in dependency of the matrix dimension: (3 \(\times \) 3: \(0 < X < 0.0049\); 4 \(\times \) 4: \(0 < X < 0.0529\); 5 \(\times \) 5: \(0 < X < 0.1369\)) |
Step 4: Aggregation of group priorities |
\(P^{\textit{LFA}}\left( {A_i } \right) =\mathop \sum \nolimits _{k=1}^K F\left( {X_j } \right) \cdot P\left( {A_{i,j} } \right) \cdot P\left( {C_j } \right) \) where \(P\left( {A_{i,j} } \right) \) is the priority of alternative \(A_i \) with respect to criterion \(C_j \), and \(P\left( {C_j } \right) \) is the priority of criterion \(C_j \)
|
Calculation of the priority vector \(\vec {v}^{{diff}}\) for representation of the preferential differences | |
---|---|
Step 1: Determination of the preferential differences \(\theta _{rij}\) for every pair of alternatives for every decision maker |
\({\theta _{rij}} = \left| {P_i^{gl}\left( {D{M_r}} \right) - P_j^{gl}\left( {D{M_r}} \right) } \right| \)
|
Step 2: Calculation of the relative importance \(b_{rij}\) comparing \(A_i\) and \(A_j\) for \(\textit{DM}_r\)
|
\({b_{rij}} = \vec v_{rk}^t\vec a_{i \cdot j}^r,\) where the vector \(\vec v_{rk}^t\) contains the decision weights about the m criteria \(C_k\) with \(\vec v_{rk}^t = \left[ {P_{{C_1}}^{lo}\left( {D{M_r}} \right) , \ldots ,P_{{C_m}}^{lo}\left( {D{M_r}} \right) } \right] \) and \(\vec {a}_{i\cdot j}^r\) is the vector of the pairwise comparison judgments comparing \(A_i\) and \(A_j\) for \(\textit{DM}_r\) with respect to each criterion |
Step 3: Determination of the aggregation matrix of pairwise comparison for all decision makers with the consideration of preferential differences \(\hbox {B}^\mathrm{diff}\)
|
\(\left[ {\begin{array}{llll} 1&{}\quad {\root \sum \nolimits _{r = 1}^R {{\theta _{r12}}} \of {{\prod \nolimits _{r = 1}^R {b_{r12}^{{\theta _{r12}}}} }}}&{}\quad \cdots &{}\quad {\root \sum \nolimits _{r = 1}^R {{\theta _{r1n}}} \of {{\prod \nolimits _{r = 1}^R {b_{r1n}^{{\theta _{r1n}}}} }}}\\ {\root \sum \nolimits _{r = 1}^R {{\theta _{r21}}} \of {{\prod \nolimits _{r = 1}^R {b_{r21}^{{\theta _{r21}}}} }}}&{}\quad 1&{}\quad \cdots &{}\quad {\root \sum \nolimits _{r = 1}^R {{\theta _{r2n}}} \of {{\prod \nolimits _{r = 1}^R {b_{r2n}^{{\theta _{r2n}}}} }}}\\ \vdots &{}\quad \vdots &{}\quad \ddots &{}\quad \vdots \\ {\root \sum \nolimits _{r = 1}^R {{\theta _{rn1}}} \of {{\prod \nolimits _{r = 1}^R {b_{rn1}^{{\theta _{rn1}}}} }}}&{}\quad {\root \sum \nolimits _{r = 1}^R {{\theta _{rn2}}} \of {{\prod \nolimits _{r = 1}^R {b_{rn2}^{{\theta _{rn2}}}} }}}&{}\quad \cdots &{}\quad 1 \end{array}} \right] \)
|
Step 4: Calculation of the priority vector \(\vec {v}^{{diff}}\) by employing the approach of normalization of the geometric mean of the rows |
\({\left( {{{\vec v}^{diff}}} \right) ^t} = \left[ {v_1^{diff}, \ldots ,v_i^{diff}, \ldots ,v_n^{diff}} \right] ,\) where \(v_i^{diff} = \frac{{{{\left( {\prod \nolimits _{j = 1}^n {b_{ij}^{diff}} } \right) }^{1/n}}}}{{\sum \nolimits _{i = 1}^n {{{\left( {\prod \nolimits _{j = 1}^n {b_{ij}^{diff}} } \right) }^{1/n}}} }}\) is the decision weight of the ith alternative |
Calculation of the priority vector \(\vec {\hbox {v}}^\mathrm{rank}\) for representation of the preferential ranks | |
---|---|
Step 1: Determination of the rank adjusting factor \(\partial _i\)
|
\(\partial _\mathrm{i}\sum _\mathrm{r=1}^\mathrm{N}\,\partial _\mathrm{ri}\) with \(\partial _{ri}=\frac{n}{\rho _{ri}}\), where \(\rho _{ri}\) denotes the preferential rank of \(A_i\), given by \(\textit{DM}_r\)
|
Step 2: Normalizing the rank adjusting factor to obtain the priority vector \(\vec {v}^{rank}\)
|
\({\left( {{{\vec v}^{rank}}} \right) ^t} = \left[ {v_1^{rank}, \ldots ,v_i^{rank}, \ldots ,v_n^{rank}} \right] ,\) whereby \(\hbox {v}_\mathrm{i}^\mathrm{rank}=\partial _\mathrm{i}/\sum _{i=1}^{n}\partial _\mathrm{i}\)
|
Calculation of overall priority vector \(P(A_i)\) for considering preferential differences and preferential ranks | |
---|---|
\(P(A_{i})=\vec {v}^{diff}\times \vec {v}^{rank}\)
|
Information\(\backslash \)aggregation techniques | AIJ (WAMM) | AIJ (WGMM) | AIP (WAMM) | AIP (WGMM) | LFA (WAMM) | LFA (WGMM) | Group AHP model |
---|---|---|---|---|---|---|---|
Pairwise comparison judgments |
\(+\)
|
\(+\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(+\)
|
Priorities |
\(-\)
|
\(-\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
Preferential differences |
\(-\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(+\)
|
Rankings |
\(-\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(+\)
|
Consistency ratio |
\(-\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(+\)
|
\(+\)
|
\(-\)
|
4.3.2 Evaluation
\(\hbox {A}_1\)
|
\(\hbox {A}_2\)
|
\(\hbox {A}_3\)
|
\(\hbox {A}_4\)
| Pareto Optimality | |||
---|---|---|---|---|---|---|---|
Individual preference values
| |||||||
\(\hbox {DM}_1\)
| Test scenario 1 | 0.236 | 0.418 | 0.164 | 0.181 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| |
Test scenario 2 | 0.161 | 0.421 | 0.159 | 0.259 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| ||
\(\hbox {DM}_{2}\)
| Test scenario 1 | 0.490 | 0.127 | 0.173 | 0.210 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| |
Test scenario 2 | 0.485 | 0.140 | 0.204 | 0.170 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| ||
\(\hbox {DM}_{3}\)
| Test scenario 1 | 0.238 | 0.262 | 0.063 | 0.437 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| |
Test scenario 2 | 0.190 | 0.248 | 0.104 | 0.458 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| ||
Group preference values
| |||||||
...satisfied ? | |||||||
AIJ (WAMM) | Test scenario 1 | 0.290 | 0.321 | 0.133 | 0.254 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| Yes |
Test scenario 2 | 0.181 | 0.353 | 0.184 | 0.281 |
\(\hbox {A}_1\,<\,\hbox {A}_3\)
| No | |
AIJ (WGMM) | Test scenario 1 | 0.279 | 0.344 | 0.13 | 0.246 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| Yes |
Test scenario 2 | 0.164 | 0.370 | 0.19 | 0.276 |
\(\hbox {A}_1\,<\,\hbox {A}_3\)
| No | |
AIP (WAMM) | Test scenario 1 | 0.312 | 0.300 | 0.146 | 0.241 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| Yes |
Test scenario 2 | 0.264 | 0.302 | 0.161 | 0.272 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| Yes | |
AIP (WGMM) | Test scenario 1 | 0.318 | 0.288 | 0.149 | 0.244 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| Yes |
Test scenario 2 | 0.252 | 0.297 | 0.172 | 0.279 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| Yes | |
LFA (WAMM) | Test scenario 1 | 0.253 | 0.343 | 0.128 | 0.275 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| Yes |
Test scenario 2 | 0.177 | 0.37 | 0.165 | 0.288 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| Yes | |
LFA (WGMM) | Test scenario 1 | 0.289 | 0.343 | 0.133 | 0.234 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| Yes |
Test scenario 2 | 0.165 | 0.375 | 0.190 | 0.269 |
\(\hbox {A}_1\,<\,\hbox {A}_3\)
| No | |
Group AHP model | Test scenario 1 | 0.370 | 0.264 | 0.041 | 0.325 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| Yes |
Test scenario 2 | 0.329 | 0.276 | 0.063 | 0.331 |
\(\hbox {A}_1\,>\,\hbox {A}_3\)
| Yes |
\(\hbox {A}_1\)
|
\(\hbox {A}_2\)
|
\(\hbox {A}_{3}\)
|
\(\hbox {A}_4\)
| Homogeneity | |||
---|---|---|---|---|---|---|---|
Individual preference values
| |||||||
\(\hbox {DM}_1\)
| Test scenario 3 | 0.332 | 0.338 | 0.166 | 0.164 |
\({\upmu }_{1.2} =2.00\)
| |
Test scenario 4 | 0.264 | 0.364 | 0.143 | 0.229 |
\({\upmu }_{4.3} =1.60\)
| ||
\(\hbox {DM}_{2}\)
| Test scenario 3 | 0.458 | 0.119 | 0.229 | 0.194 |
\({\upmu }_{1.2} =2.00\)
| |
Test scenario 4 | 0.445 | 0.139 | 0.160 | 0.257 |
\({\upmu }_{4.3} = 1.60\)
| ||
\(\hbox {DM}_{3}\)
| Test scenario 3 | 0.205 | 0.257 | 0.102 | 0.436 |
\({\upmu }_{1.2} =2.00\)
| |
Test scenario 4 | 0.232 | 0.290 | 0.183 | 0.295 |
\({\upmu }_{4.3} = 1.60\)
| ||
Group preference values
| |||||||
...satisfied? | |||||||
AIJ (WAMM) | Test scenario 3 | 0.290 | 0.271 | 0.212 | 0.228 |
\({\upmu }_{1.2} =1.37\)
| No |
Test scenario 4 | 0.268 | 0.304 | 0.149 | 0.279 |
\({\upmu }_{4.3} =1.88\)
| No | |
AIJ (WGMM) | Test scenario 3 | 0.302 | 0.311 | 0.171 | 0.216 |
\({\upmu }_{1.2} =1.76\)
| No |
Test scenario 4 | 0.268 | 0.320 | 0.145 | 0.267 |
\({\upmu }_{4.3} =1.84\)
| No | |
AIP (WAMM) | Test scenario 3 | 0.344 | 0.256 | 0.172 | 0.228 |
\({\upmu }_{1.2}=2.00\)
| Yes |
Test scenario 4 | 0.312 | 0.282 | 0.156 | 0.250 |
\({\upmu }_{4.3}=1.60\)
| Yes | |
AIP (WGMM) | Test scenario 3 | 0.352 | 0.248 | 0.176 | 0.223 |
\({\upmu }_{1.2} =2.00\)
| Yes |
Test scenario 4 | 0.312 | 0.270 | 0.161 | 0.258 |
\({\upmu }_{4.3} =1.60\)
| Yes | |
LFA (WAMM) | Test scenario 3 | 0.273 | 0.293 | 0.163 | 0.271 |
\({\upmu }_{1.2} =1.68\)
| No |
Test scenario 4 | 0.243 | 0.331 | 0.139 | 0.286 |
\({\upmu }_{4.3} =2.05\)
| No | |
LFA (WGMM) | Test scenario 3 | 0.287 | 0.322 | 0.171 | 0.220 |
\({\upmu }_{1.2} =1.68\)
| No |
Test scenario 4 | 0.266 | 0.330 | 0.146 | 0.259 |
\({\upmu }_{4.3} =1.77\)
| No | |
Group AHP model | Test scenario 3 | 0.352 | 0.238 | 0.073 | 0.337 |
\({\upmu }_{1.2}= 4.83\)
| No |
Test scenario 4 | 0.371 | 0.267 | 0.074 | 0.288 |
\({\upmu }_{4.3} =3.91\)
| No |
Synthesized value of the reciprocal of the individual assessments | Normalised | Reciprocal of the synthesized value of the original assessments | Normalised | Reciprocal property... ...satisfied? | ||
---|---|---|---|---|---|---|
Group preference values (based on test scenario 1)
| ||||||
AIP (WAMM) |
\(\hbox {A}_1\)
| 3.567 | 0.174 | 3.197 | 0.183 | No |
\(\hbox {A}_2\)
| 4.323 | 0.211 | 3.337 | 0.191 | ||
\(\hbox {A}_3\)
| 7.969 | 0.389 | 6.828 | 0.390 | ||
\(\hbox {A}_4\)
| 4.644 | 0.226 | 4.147 | 0.237 | ||
AIP (WGMM) |
\(\hbox {A}_1\)
| 3.395 | 0.180 | 3.395 | 0.180 | Yes |
\(\hbox {A}_{2}\)
| 3.754 | 0.199 | 3.754 | 0.199 | ||
\(\hbox {A}_{3}\)
| 7.271 | 0.386 | 7.271 | 0.386 | ||
\(\hbox {A}_{4}\)
| 4.425 | 0.235 | 4.425 | 0.235 | ||
LFA (WAMM) |
\(\hbox {A}_1\)
| 302.418 | 0.200 | 3.942 | 0.215 | No |
\(\hbox {A}_{2}\)
| 236.217 | 0.156 | 2.9177 | 0.159 | ||
\(\hbox {A}_{3}\)
| 469.075 | 0.310 | 7.8032 | 0.427 | ||
\(\hbox {A}_{4}\)
| 507.758 | 0.335 | 3.6310 | 0.198 | ||
LFA (WGMM) |
\(\hbox {A}_1\)
| 0.173 | 0.216 | 3.4549 | 0.190 | No |
\(\hbox {A}_{2}\)
| 0.123 | 0.154 | 2.9150 | 0.161 | ||
\(\hbox {A}_{3}\)
| 0.238 | 0.298 | 7.5168 | 0.414 | ||
\(\hbox {A}_{4}\)
| 0.265 | 0.332 | 4.265 | 0.235 | ||
Group AHP model |
\(\hbox {A}_1\)
| 0.041 | 0.160 | 2.702 | 0.080 | No |
\(\hbox {A}_{2}\)
| 0.037 | 0.144 | 3.789 | 0.112 | ||
\(\hbox {A}_{3}\)
| 0.126 | 0.490 | 24.247 | 0.717 | ||
\(\hbox {A}_{4}\)
| 0.053 | 0.205 | 3.080 | 0.091 |
5 Results and Discussion
Evaluation criteria/aggregation techniques | AIJ (WAMM) | AIJ (WGMM) | AIP (WAMM) | AIP (WGMM) | LFA (WAMM) | LFA (WGMM) | Group AHP model |
---|---|---|---|---|---|---|---|
Decision contexts
| |||||||
Group size: small (\(\sim \)2-5 Persons) |
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
Group size: large (\(>\)5 Persons) |
\(\bullet \)
|
\(\bullet \)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
Situation: common objectives |
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
Situation: divergent objectives |
\(\bullet \)
|
\(\bullet \)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
Situation: conflicting objectives |
\(-\)
|
\(-\)
|
\(+\)
|
\(+\)
|
\(\bullet \)
|
\(\bullet \)
|
\(\bullet \)
|
Consistency
| |||||||
Improvement |
\(-\)
|
\(+\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(-\)
|
Consideration |
\(-\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(+\)
|
\(+\)
|
\(-\)
|
Social choice axioms
| |||||||
Universal domain |
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
Pareto Optimality |
\(\bullet \)
|
\(\bullet \)
|
\(+\)
|
\(+\)
|
\(\bullet \)
|
\(\bullet \)
|
\(\bullet \)
|
Independence of irrelevant alternatives |
\(\bullet \)
|
\(\bullet \)
|
\(\bullet \)
|
\(\bullet \)
|
\(\bullet \)
|
\(\bullet \)
|
\(\bullet \)
|
Non-dictatorship |
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
Separability condition |
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
\(+\)
|
Unanimity condition |
\(\bullet \)
|
\(\bullet \)
|
\(+\)
|
\(+\)
|
\(\bullet \)
|
\(\bullet \)
|
\(\bullet \)
|
Homogeneity condition |
\(-\)
|
\(-\)
|
\(+\)
|
\(+\)
|
\(-\)
|
\(-\)
|
\(-\)
|
Power conditions (reciprocal property) |
\(-\)
|
\(+\)
|
\(-\)
|
\(+\)
|
\(-\)
|
\(-\)
|
\(-\)
|
\(\hbox {A}_1\)
|
\(\hbox {A}_2\)
|
\(\hbox {A}_3\)
|
\(\hbox {A}_4\)
| Ranking | Best alternative | |
---|---|---|---|---|---|---|
Group priorities (test scenario 1)
| ||||||
AIJ (WAMM) | 0.291 | 0.322 | 0.133 | 0.254 |
\(\hbox {A}_{2}\,>\,\hbox {A}_{1}\,>\,\hbox {A}_{4} \,>\,\hbox {A}_{3}\)
|
\(\hbox {A}_2\)
|
AIJ (WGMM) | 0.280 | 0.344 | 0.130 | 0.246 |
\(\hbox {A}_2\,>\,\hbox {A}_1\,>\,\hbox {A}_4\,>\,\hbox {A}_3\)
|
\(\hbox {A}_2\)
|
AIP (WAMM) | 0.313 | 0.300 | 0.146 | 0.241 |
\(\hbox {A}_{1} \, >\,\hbox {A}_{2} \, >\,\hbox {A}_{4}\, >\,\hbox {A}_{3}\)
|
\(\hbox {A}_1\)
|
AIP (WGMM) | 0.319 | 0.288 | 0.149 | 0.244 |
\(\hbox {A}_{1}\,>\,\hbox {A}_{2}\,>\,\hbox {A}_{4}\,>\,\hbox {A}_{3}\)
|
\(\hbox {A}_1\)
|
LFA (WAMM) | 0.254 | 0.343 | 0.128 | 0.275 |
\(\hbox {A}_2\,>\,\hbox {A}_1\,>\,\hbox {A}_4\,>\,\hbox {A}_3\)
|
\(\hbox {A}_2\)
|
LFA (WGMM) | 0.289 | 0.343 | 0.133 | 0.234 |
\(\hbox {A}_2\,>\,\hbox {A}_1\,>\,\hbox {A}_4\,>\,\hbox {A}_3\)
|
\(\hbox {A}_2\)
|
Group AHP model | 0.370 | 0.264 | 0.041 | 0.325 |
\(\hbox {A}_1\,>\hbox {A}_2\,>\hbox { A}_4\,>\,\hbox {A}_3\)
|
\(\hbox {A}_1\)
|