The process by which agreement (or near agreement) is reached is crucial
in consensus making. There are two approaches to achieving consensus, one
that is commonly used but fundamentally flawed, and the other that is
lesser known yet more in line with the values of a learning organization.
First the common approach. This usually involves some facilitated
discussion among a group of people with an interest in a particular
process or its outcomes. Various improvement ideas are floated and,
eventually, one is select that all can at least "live with." The problem
here is that it's a popularity contest or a power play for either the
ideas themselves or the people who advocate them (or perhaps both). Quite
often it comes down to two competing ideas (or people) and so the lowest
common denominator of both is selected. Again, can you live with it?
Another principle problem with this approach is that it doesn't
necessarily lead to the best solution possible. In fact, often there is
no objective way to even know if it is even better than what was in place
before.
Building consensus in this way is centred more on ensuring that people
are polite and respectful. This is of course a good thing. But this
approach alone is disconnected from the work itself.
The lesser known but more effective approach involves study,
experimentation and discipline (the PDSA cycle). There are any number of
scenarios that I could describe here, but I'll just describe two for sake
of brevity.
Plan
- First the process is monitored using appropriate statistical methods
(control charts)
- the objective for improvement is determined and documented
- an operational definition of improvement is determined. This might take
the form of what the control chart will look like when improvement has
been realized. This is a crucial step because without it, this could give
rise to more conflict (is everyone working to the right goal? What is the
right goal?
- an inventory of ideas is built where everyone with an interest in the
process contributes (perhaps through brainstorming).
- The inventory is reviewed collectively and affinities are identified
- the group selects the most likely prospect - if there is not agreement
over which one to try first, no problem, if it doesn't work sufficiently,
there will be an opportunity to try another approach. If it does work,
still the group may well decide to pursue another approach to see if even
better results are possible
- The experiment is developed
DO
- the experiment is conducted
Study
- what happend? What happened that was expected to happen? What didn't
happen that was expected to happen? What happened that wasn't expected to
happen? Have a look at the control chart - was variation reduced? Is it
in better control? Is the process targeted more desirably?
Act
- institute the changes and/or take steps to go through the PDSA again.
Throughout the PDSA process, lessons learned in the previous experiment
are incorporated into subsequent ones. In other words the inventory of
ideas is constantly added to and refined as lessons are learned.
The second scenario is a variant of the above but instead of doing
experiments one-at-a-time, a concept known as Design of Experiments
allows for several experiments to be consolidated in a short number of
pre-determined runs. For example, with an objective to reduce errors in
an order-entry process, one team member may feel that it's important to
have the sales people enter their orders from the field rather than send
the orders on paper to a centralized order-entry function. Another may
suggest that experienced and product-knowledgeable people should be
entering the orders. These scenarios could be tested as part of a
designed experiment along with several other factors. Much could be
learned about what effect product knowledge has on error rates and if any
other factors influence the "product knowledge" factor (say, time on the
job).
In a manufacturing process, various points of views concerning what
levels heat, pressure, chemical concentration, presence of a catalyst,
etc. could be experimented on simultaneously.
In both situations, a designed experiment will allow us to know
emperically which of the theories advanced by the group are useful and to
what extent.
The point here is that study, good experimental methods, a disciplined
approach (such as the PDSA cycle) together form a mechanism that advances
our knowledge and implicitly builds consensus and constructive team
dynamics.
Wayne J. Levin, M.A.Sc., P.Eng
Process Improvements, Inc.
PO Box 77506
North York, Ontario CANADA
M3H 6A7
--"Wayne J. Levin" <levin@astral.magic.ca>
Learning-org -- An Internet Dialog on Learning Organizations For info: <rkarash@karash.com> -or- <http://world.std.com/~lo/>