Decision making
Week 11
“Decision Making
Methods 3”
Dr Ramesh Vahidi [email protected]
MANG6312©RameshVahidi2022 1
Southampton Business School
12th May 22
Aims of week 11: Let’s Decide or Find Resolutions
• Introducing the most used Multiple Criteria
Decision Making Method, its applications,
process, potential advantages and disadvantages.
MANG6312©RameshVahidi2022 2
PART 1
Background to the AHP
method
MANG6312©RameshVahidi2022 3
Southampton Business School
MANG6312©RameshVahidi2022 4
MANG6312©RameshVahidi2022 5
“Problem Formulation,
Decision Making/Analysis
Method based on:
The Analytic Hierarchy Process (AHP)
MANG6312©RameshVahidi2022 6
Southampton Business School
What is AHP?
✓ An MCDM developed in 1970s by Prof Thomas Saaty.
✓ Based on mathematics and psychology of DM.
✓ Most used DM method with unlimited applications.
✓ Facilitates problem structuring (modelling) and making
complex decisions.
✓ Based on pairwise comparisons.
✓ NASA, HP, IBM, KPMG, Ford, Shell, governments, …
✓ Well supported by software.
✓ Widely studied, developed … still room in PM.
MANG6312©RameshVahidi2022 7
http://www.youtube.com/watch?v=vC9Aeu06wjA
PART 2
Problem structuring and
AHP Process
MANG6312©RameshVahidi2022 8
Southampton Business School
Overview of AHP: Problem Formulation
MANG6312©RameshVahidi2022 9
AHP: Problem/Decision Formulation (Structure)
MANG6312©RameshVahidi2022
Goal / Objective
Criterion 1 Criterion 2 Criterion 3 Criterion N
Alternative 1 Alternative 2 … Alternative N
10
Both terms ‘Criteria’ or ‘Attribute’ might be used in AHP. Both terms ‘Alternative’ or ‘Option’ might be used in AHP.
Extension of the Hierarchy
MANG6312©RameshVahidi2022 11
You Might also have different sub-Objectives or sub-Goals before the
criteria layer
Example – Hierarchy for Decision on Buying a Car
MANG6312©RameshVahidi2022 AHP
Buying a Car
Style Reliability Fuel Consumption
Civic Saturn Escort
12
AHP Steps (Stages)
MANG6312©RameshVahidi2022 13
Step 1: Set up the decision Hierarchy.
Step 2: Make pairwise comparisons of criteria (attributes) and alternatives.
Step 3: Transform the comparisons into weights.
Step 4: Use the weights to obtain scores for the different options and make a decision.
Step 5: Perform sensitivity analysis. (Goodwin and Wright, 2014, pp.76-80)
PART 3-1
An example of DM with
AHP
MANG6312©RameshVahidi2022 14
Southampton Business School
AHP Example
Buying a Car
MANG6312©RameshVahidi2022 15
Step 1 – Build the Hierarchy
MANG6312©RameshVahidi2022 AHP
Buying a Car
Style Reliability Fuel Consumption
Civic Saturn Escort
16
Step 1 – Build the Hierarchy
MANG6312©RameshVahidi2022 17
Buying a Car
For the next lecture, let’s decide about the car (or something else) you would like
to buy after your first few handsome monthly salaries!
Let’s Model your decision:
1) What is your goal? (be as specific as you could) → Add this to the model!
2) List your criteria for this purchase. Select the 3 main criteria (for the purposes of this exercise)? → Add these to the model.
3) List your options. Select the 3 most desired/possible options (for the purposes of this exercise)? → Add these to the model!
Try modelling a decision yourself!
Step 2 – Pairwise Comparison
Make tables like these for your decision and bring to the next lecture
MANG6312©RameshVahidi2022 18
Step 2 – Pairwise Comparisons
We need to know/ask for the
priorities through comparing them
in pairs, i.e. finding out:
• The relative importance of the
criteria for achieving the goal;
• The relative importance/merit
of the alternatives (options) in
terms of or in relation to each
criteria. MANG6312©RameshVahidi2022 19
Step 2 – Pairwise Comparison
Compare the importance of the criteria against each other
MANG6312©RameshVahidi2022 20
MANG6312©RameshVahidi2022 21
AHP Part 1
https://www.youtube.com/watch?v=ht0Wm3 3xEno
WATCH this video and see how far you could go with completing the scoring tables!
Example of AHP Steps
MANG6312©RameshVahidi2022 22
Can You Compare ‘Apples’ and ‘Oranges’?
MANG6312©RameshVahidi2022 23
Intensity of
Importance Definition Explanation
1 Equal Importance Two activities contribute equally to the
objectives
3 Moderate Importance Experience and judgement slightly favour one
activity over another
5 Strong Importance Experience and judgement strongly favour one
activity over another
7
Very strong Importance An activity is favoured very strongly over
another; its dominance demonstrated in
practice
9
Extreme Importance The evidence favouring one activity over
another is of the highest possible order of
affirmation
2, 4, 6, 8 Intermediate values
Step 2 – Basis for Pairwise Comparisons
Step 2 – Pairwise Comparison - Criteria
Compare the importance of the criteria against each other
MANG6312©RameshVahidi2022 24
Implementation Method: asking relevant decision
makers about their preferences.
MANG6312©RameshVahidi2022 25
Step 2 – Pairwise Comparison - Alternatives
Comparing Alternatives against each Criteria
Implementation Method: asking relevant decision
makers about their preferences.
Step 3 - Synthesis : Prioritizing Criteria & Alts
The process of calculating the priority of each criterion in terms of
its contribution to the overall goal:
1) Sum values in each column of pairwise comparison matrix
2) Divide each element by its column total (gives normalized
pairwise comparison matrix)
3) Compute average of elements in each row (gives estimate of
relative priorities of elements being compared)
MANG6312©RameshVahidi2022 26
MANG6312©RameshVahidi2022 27
Step 3 – Prioritize Criteria
1) Sum values in each column
2) Divide each element
by its column total
(gives normalized
pairwise comparison
matrix)
Priorities of criteria
Step 3 – Prioritize Alts in terms of Style
MANG6312©RameshVahidi2022
Alternatives Civic Saturn Escort
Civic 1 4 1/3
Saturn ¼ 1 1/7
Escort 3 7 1
SUM 4.25 12 1.47
28
STYLE
Step 3 – Prioritize Alts in terms of Reliability
Reliability
Alt.s Civic Saturn Escort
Civic 1 1/6 1
Saturn 6 1 6
Escort 1 1/6 1
SUM 8 8/6 8
Alt.s Civic Saturn Escort Avg.
Civic .125 .125 .125 .125
Saturn .75 .75 .75 .75
Escort .125 .125 .125 .125
Reliability
MANG6312©RameshVahidi2022 29
Step 3 - Prioritize Alts in terms of Fuel Cons.
Alt.s Civic Saturn Escort Avg.
Civic .755 .790 .667 .737
Saturn .151 .158 .25 .186
Escort .094 .053 .083 .077
Fuel Consumption
MANG6312©RameshVahidi2022 30
Criteria Style
.701
Reliability
.100
Fuel Cons.
.197 Final Priority
Civic .266 .125 .737 0.344
Saturn .080 .75 .186 0.168
Escort .654 .125 .077 0.486
Example:
Civic: .701 x .266 + .100 x .125 + .197 x .737 = .0344
Step 4 - Prioritize Alts based on their Total Scores
MANG6312©RameshVahidi2022 31
MANG6312©RameshVahidi2022 32
AHP Part 1
https://www.youtube.com/watch?v=ht0Wm3 3xEno
WATCH now if you had not done it already! Example of AHP Steps
PART 4
Concluding remarks
MANG6312©RameshVahidi2022 33
Southampton Business School
Review of the AHP Approach to DM
MANG6312©RameshVahidi2022 34
“When we think, we identify objects or ideas and
also relations among them. When we identify
anything, we decompose the complexity which
we encounter. When we discover the relations,
we synthesize. This is the fundamental process
underlying perception: decomposition and
synthesis.”
(Saaty, 1988, p.3)
Further Remarks …
• Consistency →
- A is 2 times preferred to B, B is 3 times preferred to C ➔ A should be 6 times preferred to C
- Higher than a percentage not accepted (→ perhaps random comparisons)
• Sensitivity analysis →
- How the final results will change with the change of comparisons
- Aided with software
• AHP in group DM →
- Comparisons discussed (our previous discussions re group DM applies)
- Average of comparisons (matter of extensive research)
MANG6312©RameshVahidi2022 35
AHP Advantages
• Structured approach: Clarity in problem formulation,
analysis process, documentation, tracing back and tracking
the DM Process;
• Multiple objectives: Possibility of considering and dealing
with conflicting objectives;
• Facilitates group DM: Different decision makers can
share their ideas and easily get involved in the procedure;
MANG6312©RameshVahidi2022 36
AHP Advantages
• No limitation for Criteria: Both tangible and
intangible criteria;
• No limitation for Measures: Both qualitative and
quantitative measures;
• Ease of application: No need for DM to know the
complex math (just sum and average of course!) use
of software. MANG6312©RameshVahidi2022 37
Readings
Saaty, T. (2008) ‘Decision making with the analytic hierarchy process’. Int. J. Services Sciences, 1(1). pp. 83-98. (http://www.rafikulislam.com/uploads/resourses/197245512559a37aadea6d.pdf)
Vaidya, O.S. & Kumar, S. (2006) ‘Analytic Hierarchy Process: An overview of applications’, European Journal of Operational Research, 169, pp.1-29.
von Solms, SH (2011) ‘Validity of the AHP/ANP: Comparing Apples and Oranges’. International Journal of the Analytic Hierarchy Process, 3(1), pp. 2-27.
W. Saaty, R. (2007) ‘Validating the Analytic Hierarchy/Network Processes’. ISAHP- International Symposium on the Analytic hierarchy process. Viña del Mar, Chile, August 3-8, pp.1-27.
Zahedi, F. (1986) ‘The Analytic Hierarchy Process - A Survey of the Method and its Applications’. Interfaces, 16 (4), pp. 96-108.
MANG6312©RameshVahidi2022 38
Readings Al-Harbi , K.M.A. (2001) ‘Application of the AHP in project management’, International Journal of Project Management, 19 (2001), pp.19-27.
Goodwin, P. & Wright, G. (2014) Decision analysis for management judgement. 5th edn. West Essex: John Wiley & Sons. (for AHP and other models)
Ho, W. (2008) ‘Integrated analytic hierarchy process and its applications - A literature review’, European Journal of Operational Research,186, pp.211–228.
Ishizaka, A. & Labib, A. (2009) ‘Analytic Hierarchy Process and Expert Choice - Benefits and Limitations’, OR Insight, 22(4), pp. 201–220.
MANG6312©RameshVahidi2022 39
40 MANG6312©RameshVahidi2022