10. Techniques
10.16 Decision Analysis
BABOK® Guide
10.16.1 Purpose
Decision analysis formally assesses a problem and possible decisions in order to determine the value of alternate outcomes under conditions of uncertainty.
10.16.2 Description
Decision analysis examines and models the possible consequences of different decisions about a given problem. A decision is the act of choosing a single course of action from several uncertain outcomes with different values. The outcome value may take different forms depending on the domain, but commonly include financial value, scoring, or a relative ranking dependent on the approach and evaluation criteria used by the business analyst.
Decisions are often difficult to assess when:
- the problem is poorly defined,
- the action leading to a desired outcome is not fully understood,
- the external factors affecting a decision are not fully understood, or
- the value of different outcomes is not understood or agreed upon by the various stakeholders and does not allow for direct comparison.
Decision analysis helps business analysts evaluate different outcome values under conditions of uncertainty or in highly complex situations. A variety of decision analysis approaches are available. The appropriate approach depends on the level of uncertainty, risk, quality of information, and available evaluation criteria. Effective decision analysis requires an understanding of:
- the values, goals, and objectives that are relevant to the decision problem,
- the nature of the decision that must be made,
- the areas of uncertainty that affect the decision, and
- the consequences of each potential decision.
Decision analysis approaches use the following activities:
- Define Problem Statement: clearly describe the decision problem to be addressed.
- Define Alternatives: identify possible propositions or courses of action.
- Evaluate Alternatives: determine a logical approach to analyze the alternatives. An agreement of evaluation criteria can also be determined at the beginning of this activity.
- Choose Alternative to Implement: the stakeholders responsible for making the decision choose which alternative will be implemented based on the decision analysis results.
- Implement Choice: implement the chosen alternative.
There are a number of decision analysis tools available to assist the business analyst and decision makers in making objective decisions. Some of the tools and techniques are best for deciding between two alternatives, while others handle multiple alternatives.
Some general decision analysis tools and techniques include:
- pro versus con considerations,
- force field analysis,
- decision tables,
- decision trees,
- comparison analysis,
- analytical hierarchy process (AHP),
- totally-partially-not (TPN),
- multi-criteria decision analysis (MCDA), and
- computer-based simulations and algorithms.
10.16.3 Elements
.1 Components of Decision Analysis
General components of decision analysis include:
- Decision to be Made or Problem Statement: a description of what the decision question or problem is about.
- Decision Maker: person or people responsible for making the final decision.
- Alternative: a possible proposition or course of action.
- Decision Criteria: evaluation criteria used to evaluate the alternatives.
.2 Decision Matrices
The tables below provide examples of a a simple decision matrix and a weighted decision matrix.
A simple decision matrix checks whether or not each alternate meets each criterion being evaluated, and then totals the number of criteria matched for each alternate. In this example, Alternate 1 would likely be selected because it matches the most criteria.
Table 10.16.1: Simple Decision Matrix
10.16.3.2.2 |
Alternate 1 |
Alternate 2 |
Alternate 3 |
|
Criterion 1 |
Meets criterion |
n/a |
n/a |
|
Criterion 2 |
Meets criterion |
Meets criterion |
Meets criterion |
|
Criterion 3 |
n/a |
Meets criterion |
Meets criterion |
|
Criterion 4 |
Meets criterion |
n/a |
n/a |
|
Score |
3 |
2 |
2 |
A weighted decision matrix assesses options in which each criterion is weighted based on importance. The higher the weighting, the more important the criterion. In this example, the criteria are weighted on a scale of 1-5, where 5 indicates the most important. The alternates are ranked per criterion on a scale of1-5, where 5 indicates the best match. In this example, Alternate 3 would likely be selected due to its high weighted score.
Table 10.16.2: Weighted Decision Matrix
10.16.3.2.4 |
Criterion Weighting |
Alternate 1 |
Alt 1 Value |
Alternate 2 |
Alt 2 Value |
Alternate 3 |
Alt 3 Value |
|
Criterion 1 |
1 |
Rank = 1*3 |
3 |
Rank = 1*5 |
5 |
Rank = 1*2 |
2 |
|
Criterion 2 |
1 |
Rank = 1*5 |
5 |
Rank = 1*4 |
4 |
Rank = 1*3 |
8 |
|
Criterion 3 |
3 |
Rank = 3*5 |
15 |
Rank = 3*1 |
3 |
Rank = 3*5 |
15 |
|
Criterion 4 |
5 |
Rank = 5*1 |
5 |
Rank = 5*5 |
25 |
Rank = 5*3 |
15 |
|
Weighted Score |
|
|
28 |
|
37 |
|
40 |
For more information on decision trees, see Decision Modelling (p. 265).
.3 Decision Trees
A decision tree is a method of assessing the preferred outcome where multiple sources of uncertainty may exist. A decision tree allows for assessment of responses to uncertainty to be factored across multiple strategies.
Decision trees include:
- Decision Nodes: that include different strategies.
- Chance Nodes: that define uncertain outcomes.
- Terminator or End Nodes: that identify a final outcome of the tree.
.4 Trade-offs
Trade-offs become relevant whenever a decision problem involves multiple, possibly conflicting, objectives. Because more than one objective is relevant, it is not sufficient to simply find the maximum value for one variable (such as the financial benefit for the organization). When making trade-offs, effective methods include:
- Elimination of dominated alternatives: a dominated alternative is any option that is clearly inferior to some other option. If an option is equal to or worse than some other option when rated against the objectives, the other option can be said to dominate it. In some cases, an option may also be dominated if it only offers very small advantages but has significant disadvantages.
- Ranking objectives on a similar scale: one method of converting rankings to a similar scale is proportional scoring. Using this method, the best outcome is assigned a rating of 100, the worst a rating of 0, and all other outcomes are given a rating based on where they fall between those two scores. If the outcomes are then assigned weights based on their relative importance, a score can be assigned to each outcome and the best alternative assigned using a decision tree.
10.16.4 Usage Considerations
.1 Strengths
- Provides business analysts with a prescriptive approach for determining alternate options, especially in complex or uncertain situations.
- Helps stakeholders who are under pressure to assess options based on criteria, thus reducing decisions based on descriptive information and emotions.
- Requires stakeholders to honestly assess the importance they place on different alternate outcomes in order to help avoid false assumptions.
- Enables business analysts to construct appropriate metrics or introduce relative rankings for outcome evaluation in order to directly compare both the financial and non-financial outcome evaluation criteria.
.2 Limitations
- The information to conduct proper decision analysis may not be available in time to make the decision.
- Many decisions must be made immediately, without the luxury of employing a formal or even informal decision analysis process.
- The decision maker must provide input to the process and understand the assumptions and model limitations. Otherwise, they may perceive the results provided by the business analyst as more certain than they are.
- Analysis paralysis can occur when too much dependence is placed on the decision analysis and in determining probabilistic values.
Some decision analysis models require specialized knowledge (for example, mathematical knowledge in probability and strong skills with decision analysis tools).
