Quantitative and Qualitative Decision Making

Decision-Making Resources > Decision-Making Tools & Techniques > Quantitative and Qualitative Decision Making

Most situations are a mix of quantitative and qualitative decision-making. Often some of the information is measurable and clearly quantitative, and some is very qualitative, a gut-feel at best. One Robust Decisions' client, a manufacturer of rocket engines, makes many early decisions qualitatively and transitions to quantitative, as the project is refined.

The challenge is that many problems are made up of a mix of qualitative and quantitative decision-making. Of course you can develop scales to make the qualitative measurable, but this is time consuming and always suspect. On the other hand, there has been much literature recently on the value of making qualitative decisions. It has been argued that gut feel is the key to good decisions (see Malcolm Gladwell's 2005 book, Blink).

Compounding this problem is then many decisions are based on the use of 5-point, textual scales. A number of common scales are listed in the table below.

	Probability	Relative goodness	Agreement	Frequency	Importance	Quality
5	Very High	much better than	Strongly agree	Always	Very important	Excellent
4	High	better than	Somewhat agree	Frequently	Important	Above average
3	Medium	about the same as	Undecided	Occasionally	Neutral	Average
2	Low	Worse than	Somewhat disagree	Rarely	Minor importance	Below average
1	Very low	much worse than	Strongly disagree	Never	Unimportant	Poor

The problem with these scales is that some of these are being proxy for qualitative information and some for quantitative.

In general, methods that support decision making either support qualitative or quantitative information exclusively. But then the challenge is to provide a method that can support these diverse types of information on which real-world decisions are based. The approach taken by Robust Decisions is to transpose all evaluations, regardless of type, into probabilities. Probabilities are exceedingly hard to estimate. Thus, Robust Decisions products provide graphical user interfaces to capture both the evaluation and the certainty.

Belief Map example For example, three team members use a Belief Map (shown right) to evaluate how well three proposals meet the requirement "easy to implement." This assessment is one of many used to choose which of the three proposals (red, green or blue) to fund. John Franks has moved his point indicating how easy the red proposal is to implement (the larger red dot) to a point signifying that it is medium and he has high certainty. The location of this point is transformed into a probability using a patented Bayesian method.

A similar approach is taken with quantitative assessments. For example, another goal of the proposal assessment is that the total time to implement be less than 20 hours. This delighted value or target is represented on the number line as a hand with the thumb up. However, all goals are soft, and thus it is admitted, from the beginning that a proposal that is good in other areas may take up to 30 hours. Any time longer than 30 hours is not acceptable. This disgusted value, or threshold, is represented on the number line is a hand with the thumb down. The red proposal claimed that the work could be done in 20 hours. However, more careful reading left John with the impression that most likely it would take 23 hours and could even take 25 hours. These values for the high low and most likely estimates are shown on the number line. These are also transformed into a probability to be combined with that for the qualitative assessment.

All robust decision products make use of these types of techniques to support quantitative and qualitative decision-making. Download free versions of Accord products and explore how we combine the different types of information.

Free 30-day Trial: Try Accord Professional Free -- Download Here!

Learn More

Decision-Maker Blog

» view all posts