Later On

A blog written for those whose interests more or less match mine.

Making decisions in a group

with 2 comments

Groups have different processes for making decisions—some wait for consensus, others vote, and in others the leader decides. I found, long ago, a little DOS program called “Best Choice” that greatly simplified group decision-making. Using it consisted of a few simple steps:

List the possible options or choices from which you are to select.

List the criteria you will use to evaluate those.

Weight the criteria as you want—that is, assign a number to each criterion to indicate how heavily it should weigh in the final decision. (You actually can do this using the program: do a run with the criteria being the choices, using the sole criterion “Importance.” After going through the decision routine, the result will be the weights for the criteria, in the judgment of the group.)

List the people who will be evaluating the options according to the criteria.

Weight the individuals as you want—again, assign a number to each person to see how much weight to give his or her opinions. (For example, an expert in the field might be given a greater weight than someone who knows little about the matter.) One nice thing: you can change the weights of the deciders to see what effects that would have, and if you weight all but one as zero, you can see how that one person ranked the choices.

The program then presents each decider with a set of pairs of the options for each criterion. Each decider than selects which of each pair is “better” given the criterion being considered.

Note the simplification: instead of considering the whole range of choices and criteria, the decision becomes a series of small decisions between two choices using a single criterion. These decisions are easily and quickly made.

The program then uses those choices and the weights (of criteria and of decision makers) to rank the choices, showing the “value” of each option. Sometimes a group of options will have values that are close—more or less tied—and sometimes options will have values that are far apart.

It works quite well, and now there’s a Windows version available. You can view a demo of it here.

One example: I was leading a major software project, and I wanted to minimize the risks. So I brought the team together for a brainstorming session: “Assume the project has failed. What problem was the cause of failure?” We produced a list of possible problems. I then used the program and listed the problems and two criteria: How likely is the problem to happen, and how big an impact would the problem have if it did happen.

Each team member then went through the random pairings of problems, first evaluating each pair and indicating which one of each pair was more likely to happen, and then going through another set of pairs indicating which one of each pair would have a greater impact if it did happen. The program then ranked the problems based on the input of the entire team (appropriately weighted) and we had our risk factors identified in terms of their danger.

I’ve also used it to pick vacation spots, cat names, cars, and so on.

UPDATE: Link fixed.

Written by Leisureguy

22 August 2008 at 3:20 pm

2 Responses

Subscribe to comments with RSS.

  1. How did it work with the cats’ names? I didn’t realize there was a lot of risk associated with that …. unless maybe the cat didn’t like the name.



    23 August 2008 at 10:12 pm

  2. I don’t understand your mention of “risk”—this post describes a program that facilitates combining a group’s opinions about a set of choices measured by a set of criteria. It can be used for risk assessment (the example I discussed), but it can also be used for low-risk decisions.

    For example, suppose 4 people are thinking about going someplace for dinner, and they come up with 5 possible restaurants. The criteria they discuss are travel time, cost, ambiance, and food quality. The program just makes it very easy to combine the total set of opinions, with each person having to decide only between two restaurants at a time, and that with respect to one criterion. Once all those simple, easy decisions are made, the program rolls up all the input and presents a ranked list.

    In the case of the cat’s name, the idea was to select the name that was most appealing to the entire family. Again, it was easy to list the set of names under consideration, and then run through a selection quickly with respect to one criterion. No risk, but also no argument: the program produced the names in rank order of family appeal.



    24 August 2008 at 7:43 am

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: