Benjamin writes:
Julie wisely states:
> One of the disciplines I put on myself around consensus
> decisions is never to say - well I was right after all.
I couldn't agree more! I've heard people say more than
once, "I'll go along with the group, but I reserve 'I told you
so' rights." That dog don't hunt!
People make mistakes; teams make mistakes. I don't see
a need to embarrass a person or a group when they make
a mistake. . .everyone does it.
--- end of quotes ---
Well - yes and no. The problem with LEAVING IT at 'people make mistakes'
is that this risks leads the team to failing to act on one of its key
learning process - that of collective reflection on errors.
The work of Jim Reason ("Human Error") tells us that there are a whole
range of systemic processes which increase or decrease the probability of
a mistake happening, and that these probabilities are measureable and
their frequency (although not their actual occurances) is predictable. A
core team process must be to reflect on mistakes in order to surface the
systemic latent factors which led to them, thereby making adjustments to
the process and reducing the probability that the mistake will recur.
The opportunities for improvement in such a process are spectacular.
Whereas the probability of an error occurring during the commission of a
routine and familiar task being undertaken by a motivated and well trained
person is 0.0005, and the probability of one occurring during the
commission of an unfamiliar or novel task can be as high as 0.075,
dysfunctional systemic factors can multiply these figures by up to x17.
The interesting thing is that the PERSONAL factors which we think of as
error producing - for example boredom or tiredness - are relatively low
multipliers (X2 to X4), whereas systemic factors are large multipliers.
For example 'unclear organisational goals' is a x11 multiplier,
Phillip Capper
Wellington
New Zealand
-- pcapper@actrix.gen.nz
Learning-org -- An Internet Dialog on Learning Organizations For info: <rkarash@karash.com> -or- <http://world.std.com/~lo/>