The Psychology of Quality and More

Voting variants

Quality Tools > Tools of the Trade > 21: Voting variants

Making decisions in groups is probably the one thing that wastes more time in business meetings than anything else. How often have you sat with a group around a table, watching everyone (possibly including yourself) arguing and posturing, cajoling and negotiating, yet never seeming to get closer to agreement. The decision, in the end, is either taken unilaterally by a person or sub-group or by some form of consensus.

Voting is in our blood. Being brought up in a democratic political makes voting seem fair, and even if we do not always agree with the outcome, we will usually agree to abide by the final decision.

Who votes?

Usually, when a group votes on something, like the items from a brainstorming session to carry forwards for action, everyone has an equal vote. But what if some people are more expert than others? What if some have more authority? What about the people who are not there? And what of those who will have to carry out the selected action? What about the chairperson—do they have the casting vote?

A simple solution to this dilemma is to give more votes to the people with more weight. Thus an expert may be given two votes for other people's one. Managers may also demand right of veto, particularly if they are responsible for the decision or must pay for its implementation.

Before anything else, if this is likely to be a consideration (although hopefully it is not), sort it out. It can cause considerable acrimony if a manager overrides the decision or another person bypasses it to implement their own solution.

Criteria

Often, when people disagree about something, they are using different criteria. For example, when selecting a solution to implement for a problem, one person may be looking for the lowest cost solution, whilst another wants a solution that will please a particular group of people (such as customers!).

Criteria may be positive or negative. That is, they may be reasons for doing something, such as improving customer satisfaction or reducing defects, or they may reasons for not doing something, like excessive cost or effort. It can help to word criteria so that they all sound positive. It should also be very clear what they mean. For example, use 'low cost of implementation', rather than 'cost' or 'budget considerations'.

It is not always necessary to identify criteria, especially when a quick decision on an unimportant topic is being taken. A sign that criteria might be useful is when there is a lot of discussion and argument about the subject before the decision must be made.

Voting methods

There are a number of approaches that can be taken when voting, as described below. The method you choose may depend upon the importance of the topic in hand and the makeup of the group.

·       The simplest method is a quick show of hands, although this can cause privacy problems, as described above.

·       Another quick and simple method, especially with the multiple voting methods described below, is for everyone to write their own votes against the alternative solutions which have just been written up on a whiteboard or flipchart.

·       The most private method is to write your votes on anonymous voting slips, giving them to a trusted person, like an independent facilitator, to count and give you back the results.

·       You can have one vote each or a number of votes to distribute amongst the possible solutions.

·       You can vary the value of each vote. A popular method is for each person to have three votes of value 3 points, 2 points and 1 point. When these votes are cast, the favourite selection is more likely to stand out than if three votes each of value 1 point were used.

·       Taking the 3-2-1 approach one step further gives is the '£100' method. Each person is given 100 points to allocate amonst the possible solutions. They may give all 100 points to one solution or give one point each to 100 solutions.

·       A more considered approach is to use a series of elimination sessions, where either a proportion (eg. 50%) or a fixed number (eg. 10, then 3, then 1) are carried forwards each time. Between each voting session, the remaining possible solutions are discussed and compared. This takes more time, but can give a better eventual solution.

Making sure

One of the traps of voting is that because everyone has agreed, it is assumed that the solution selected is correct and the only solution to use. This is not, of course, necessarily so and it may well be prudent to test or otherwise verify the selected item before proceeding. As with other considerations, this should be done with a modicum of common sense, spending more effort testing solutions that will have a greater impact if they fail.

Next time: Decision trees (part 1)

This article first appeared in Quality World, the journal of the Institute for Quality Assurance