You are on page 1of 4

REVISED

Z01_REND6289_10_IM_MOD1.QXD 5/15/08 10:55 AM Page 269

M O D U L E

Analytic Hierarchy Process

TEACHING SUGGESTIONS
Teaching Suggestion M1.1: Using Multifactor Decision-Making Techniques. Many decisions students make involve a number of factors. Thus, multifactor decision-making techniques can be useful and practical. This section can be started by having students give examples of decisions that require the analysis of multiple factors. Buying a car or stereo and picking the best job offer are examples. Once students understand the principles of multiplying factor weights times factor evaluations, they will be able to understand the use of AHP. Teaching Suggestion M1.2: Using AHP. Have the students describe situations where AHP would be preferred over the multifactor evaluation process. You may want to take one of these situations and show how pairwise comparisons can be made. Students can then be asked to complete the AHP problem and determine the best solution. This can lead to in-class discussions on the AHP process.

Factor evaluations:
Factor Price Color Warranty Size Brand name Sun 0.7 0.9 0.8 0.8 0.9 Hitek 0.6 0.9 0.9 0.8 0.9 Surgo 0.8 0.4 0.4 0.2 0.6

Evaluation of SUN:
Factor Name Price Color Warranty Size Brand name Total Factor Rating 0.4 0.1 0.1 0.1 0.3 1.0 Factor Evaluation 0.7 0.9 0.8 0.8 0.9 Weighted Evaluation 0.28 0.09 0.08 0.08 0.27 0.80

SOLUTIONS TO QUESTIONS AND PROBLEMS


M1-1. Multifactor decision making is appropriate when a decision involves a number of factors. Deciding to buy a house, for example, can involve the price, location, taxes, utilities, and so forth. M1-2. When using multifactor decision making, each factor receives an importance weight. These weights will sum to 1. Then every alternative and factor combination will receive a factor evaluation. The factor weights are multiplied by the factor evaluations to get a weighted evaluation for each alternative. The alternative with the highest weighted evaluation is selected. M1-3. The analytic hierarchy process should be used when it is difcult or impossible to determine factor weights and factor evaluations subjectively. In this case, pairwise comparisons are performed to assist in the decision-making process and determine the best alternative. M1-4. Here is an analysis of Georges decision. Factor weights:
Factor Price Color Warranty Size Brand name Importance (Weight) 0.4 0.1 0.1 0.1 0.3

Evaluation of HITEK:
Factor Name Price Color Warranty Size Brand name Total Factor Rating 0.4 0.1 0.1 0.1 0.3 1.0 Factor Evaluation 0.6 0.9 0.9 0.8 0.9 Weighted Evaluation 0.24 0.09 0.09 0.08 0.27 0.77

Evaluation of SURGO:
Factor Name Price Color Warranty Size Brand name Total Factor Rating 0.4 0.1 0.1 0.1 0.3 1.0 Factor Evaluation 0.8 0.4 0.4 0.2 0.6 Weighted Evaluation 0.32 0.04 0.04 0.02 0.18 0.60

SUN is selected, with the highest total weighted evaluation of 0.80.

269

REVISED
Z01_REND6289_10_IM_MOD1.QXD 5/15/08 10:55 AM Page 270

270

MODULE 1

ANALYTIC HIERARCHY PROCESS

M1-5.
Price Car 1 Car 2 Car 3

Lindas problem can be analyzed as follows:


Car 1 Car 2 2 Car 3 7 4

Consistency information follows: Weighted sum vector Consistency vector Lambda Value of CI RI CR M1-8.
Factors Price Warranty Style Price Warranty 2 Style 9 6

(0.7096 2.0468 0.2460) (3.0011 3.0031 3.0004) 3.0015 0.0008 0.5800 0.0013

The following will be the priorities for price: Priority for car 1 is 0.6025. Priority for car 2 is 0.3151. Priority for car 3 is 0.0824. Consistency information follows: Weighted sum vector Consistency vector Lambda Value of CI RI CR M1-6.
Warranty Car 1 Car 2 Car 3 Car 1 Car 2
1 3

(1.8096 0.9460 0.2473) (3.0035 3.0019 3.0005) 3.0020 0.0010 0.5800 0.0017
Car 3
1 8 1 5

The following will be the priorities for the factors: Priority for price is 0.6049. Priority for warranty is 0.3337. Priority for style is 0.0614. Consistency information follows: Weighted sum vector Consistency vector Value of CI RI CR (1.8246 1.0044 0.1842) (3.0163 3.0097 3.0016) 0.0046 0.5800 0.0079

The following will be the priorities for warranty: Priority for car 1 is 0.0768. Priority for car 2 is 0.1863. Priority for car 3 is 0.7370. Consistency information follows: Weighted sum vector Consistency vector Lambda Value of CI RI CR M1-7.
Style Car 1 Car 2 Car 3 Car 1 Car 2
1 3

The following are the nal rankingsCar 1 is selected.


Item Car 1 Car 2 Car 3 Ranking 0.4045 0.2946 0.3008

(0.2310 0.5640 2.2825) (3.0088 3.0276 3.0972) 3.0445 0.0223 0.5800 0.0384
Weight A B C Cost 0.6 4 8 7 Reputation 0.2 9 5 6 Quality of life 0.2 7 7 3

M1-9. The weighted averages of these scores are shown in the table. Gina should choose Univesity B.
Weighted Average 5.6 7.2 6.0

Car 3 3 8

M1-10.

Using AHP, we have the following matrices.


A 1 5 3 9 B C 0.2 0.333333 1 3 0.333333 1 1.533333 4.333333

The following will be the priorities for style: Priority for car 1 is 0.2364. Priority for car 2 is 0.6816. Priority for car 3 is 0.0820.

Cost A B C Column Total

Normalized A B C

A 0.1111 0.5556 0.3333

B 0.1304 0.6522 0.2174

C 0.0769 0.6923 0.2308

Factor Evaluation (Row Average) 0.1062 0.6333 0.2605

REVISED
Z01_REND6289_10_IM_MOD1.QXD 5/15/08 10:55 AM Page 271

MODULE 1

ANALYTICAL HIERARCHY PROCESS

271

Reputation A A 1 B 0.142857 C 0.2 Column Total 1.342857

B 7 1 3 11

C 5 0.333333 1 6.333333 Factor Evaluation (Row Average) 0.7235 0.0833 0.1932

The following will be the priorities for price: Priority for system 1 (S-1) is 0.6039. Priority for system 2 (S-2) is 0.3258. Priority for system 3 (S-3) is 0.0703. Consistency information follows: Weighted sum vector (1.8178 0.9792 0.2109) (3.0099 3.0056 3.0011) 3.0055 0.0028 0.5800 0.0048
S-2 1 S-3 6 4

Normalized A B C

A 0.7447 0.1064 0.1489

B 0.6364 0.0909 0.2727

C 0.7895 0.0526 0.1579

Consistency vector Lambda Value of CI RI CR

Quality of Life A B C Column Total

A 1 1 0.2 2.2

B 1 1 0.142857 2.142857

C 5 7 1 13

Brand Name S-1 S-2 S-3 Factor Evaluation (Row Average) 0.4353 0.4866 0.0782 Quality of life 7 2 1 10 Factor Evaluation (Row Average) 0.6814 0.2160 0.1026

S-1

Normalized A B C Factors Cost Reputation Quality of life Column Total

A 0.4545 0.4545 0.0909

B 0.4667 0.4667 0.0667

C 0.3846 0.5385 0.0769

The following will be the priorities for brand name: Priority for system 1 (S-1) is 0.4838. Priority for system 2 (S-2) is 0.4232. Priority for system 3 (S-3) is 0.0930. Consistency information follows: Weighted sum vector (1.4649 1.2789 0.2794) (3.0278 3.0220 3.0051) 3.0183 0.0092 0.5800 0.0158
S-2
1 2

Cost 1 0.333333 0.142857 1.47619

Reputation 3 1 0.5 4.5

Consistency vector Lambda Value of CI RI CR

Normalized Cost Reputation Quality of life

Quality Cost Reputation of life 0.6774 0.6667 0.7000 0.2258 0.2222 0.2000 0.0968 0.1111 0.1000

Memory S-1 S-2 S-3

S-1

S-3
1 7 1 6

Using the factor weights, we nd the following weighted averages for each university.
Cost Reputation Quality of life 0.1062 0.7235 0.4353 0.6333 0.0833 0.4866 0.2605 0.1932 0.0782 0.6814 0.2160 0.1026 Weighted Average 0.2733 0.4995 0.2272

The following will be the priorities for memory: Priority for system 1 (S-1) is 0.0919. Priority for system 2 (S-2) is 0.1535. Priority for system 3 (S-3) is 0.7545. Consistency information follows: Weighted sum vector Consistency vector Lambda Value of CI M1-11. The analysis to determine which computer system is to be selected is as follows:
Price S-1 S-2 S-3 S-1 S-2 2 S-3 8 5 Speed S-1 S-2 S-3

A B C Weights

(0.2765 0.4631 2.3192) (3.0078 3.0164 3.0736) 3.0326 0.0163 0.5800 0.0281
S-2
1 3

Therefore, Gina should choose University B.

RI CR
S-1

S-3 2 5

REVISED
Z01_REND6289_10_IM_MOD1.QXD 5/15/08 10:55 AM Page 272

272

MODULE 1

ANALYTIC HIERARCHY PROCESS

Table for Factors for Problem M1-11


Brand Name 9 PC Compatibility 2
1 5 1 6 1 6 1 2

Factors Price Brand name Memory Speed Flexibility PC compatible

Price

Memory 4
1 2

Speed 5 1 2

Flexibility 3
1 4 1 2 1 3

The following will be the priorities for speed: Priority for system 1 (S-1) is 0.2299. Priority for system 2 (S-2) is 0.6479. Priority for system 3 (S-3) is 0.1222. Consistency information: Weighted sum vector Consistency vector Lambda Value of CI RI CR
Flexibility S-1 S-2 S-3 S-1

Consistency information follows: Weighted sum vector Consistency vector Lambda Value of CI RI CR (2.1717 0.2371 0.6218) (3.0389 3.0040 3.0122) 3.0184 0.0092 0.5800 0.0158 0.3849 0.0447 0.0816 0.0514 0.149 0.288

(0.6902 1.9485 0.3667) (3.0026 3.0071 3.0013) 3.0037 0.0018 0.5800 0.0032
S-2
1 2

The following will be the weights for the factors: Weight for price is Weight for brand name is Weight for memory is Weight for speed is Weight for exibility is Weight for PC compatibility is Consistency information follows: Weighted sum vector 2.39 0.275 0.312 0.918 0.493 1.801

S-3
1 8 1 4

See the table for factors for Problem M1-11.

The following will be the priorities for exibility: Priority for system 1 (S-1) is 0.0909. Priority for system 2 (S-2) is 0.1818. Priority for system 3 (S-3) is 0.7273. Consistency information follows: Weighted sum vector Consistency vector Lambda Value of CI RI CR
PC Compatibility S-1 S-2 S-3

Consistency vector = Value of CI RI CR 0.0288 1.2400 0.0232

6.2208 6.1480 6.0362 6.0592 6.1485 6.2518

(0.2727 0.5455 2.1818) (3.0000 3.0000 3.0000) 3.0000 0.0000 0.5800 0.0000
S-1 S-2 8 S-3 4
1 3

The following are the nal rankingssystem 1 (S-1) is selected.


Item System 1 (S-1) System 2 (S-2) System 3 (S-3) Ranking 0.4928 0.2400 0.2671

The following will be the priorities for PC compatibility: Priority for system 1 (S-1) is 0.7146. Priority for system 2 (S-2) is 0.0789. Priority for system 3 (S-3) is 0.2064.

You might also like