I have changed the name of this thread to represent the topic we are now engaged upon.
I am delighted to be able to agree with Ken. So perhaps we should state that the enemy is SQC as currently practiced, not to be confused with SQC as Deming developed it and argued for.
A strong undercurrent here, though, is yet another topic of great interest to this group: Can we have a science of design?
Wait! I know. This is a topic that has been beaten over and over again. I am as tired of the old discussions as I am tired of the discussion "What is design?". Still, although the topic has been beaten, it has not been beaten to death -- it is still alive and thriving. Moreover, some of us think we can actually make progress.
And the reason this comes up in this discussion is that many of those who preach today's version of SQC also preach their version of Design Science, which means optimization methods. Algorithmic methods. If people make errors, it means we need better rules and better training. Design? What has that got to do with it.
Aesthetics? Where does it fit in the equations? Yes, aesthetics and emotions matter, so show us how to put them into the matrices.
I keep trying to explain to these people that they are being too logical. They have to take people the way they are (the way we all are), not the way they would like people to be. Logic, I explain, is an artificial way of thinking. Otherwise, why would we have to teach it? Why would it be so difficult to learn?
Mind you, even engineers reason emotionally. First their emotion tells them the solution, then they invent a logical explanation and rationalization. So they pretend it was all done with equations. Hah. It's much more like we all do budgets. We know what answers we want, so we twiddle the numbers until it comes out right. That's how scientific design is done. Twiddle the weights on the matrix rows until the answer comes out right. Hypocrisy rules, even if it is subconscious, fooling even the person who does it that way.
But I digress.
Good to hear you agree with me, Ken.
Don
-----Original Message-----
From: PhD-Design - This list is for discussion of PhD studies and related research in Design [mailto:[log in to unmask]] On Behalf Of Ken Friedman
Sent: Monday, March 09, 2009 5:31 PM
To: [log in to unmask]
Subject: Re: Production Models at Hybrid Scales
Dear Don,
This is a dangerous downside, but it is not the downside of Deming. It's
the downside of people like the manager you worked for. This fellow
never bothered to read what Deming himself wrote on issues like these.
Your manager got Deming backwards.
This manager did not understand the core of statistical quality control.
Statistical quality control measures process outputs of a system under
way. You measure the system, not the people. This is a dramatic contrast
with post-facto performance measurements of individuals.
Your manager did not institute statistical quality control. He used
Management by Objectives and the quota system. Deming opposed both.
In point 10 of the Fourteen Points, Deming states that managers should
not set the kind of MBO targets that your manager imposed on your group.
"10. Eliminate slogans, exhortations, and targets for the work force
asking for defects and new levels of productivity. Such exhortations
only create adversarial relationships, as the bulk of the causes of low
quality and low performance belong to the system and thus lie beyond the
power of the work force."
In point 11, Deming says to eliminate production quotas and numerical
goals.
"11. Eliminate work standards and quotas on the factory floor.
Substitute leadership. Eliminate management by objective. Eliminate
management by numbers and numerical goals. Substitute leadership.”
Deming would have laughed at the notion of demanding a specific number
of scientific breakthroughs in each quarter. This is poor leadership
practice. It would have been ludicrous to Deming -- as a mathematical
physicist and working scientist, he knew exactly how difficult it is to
achieve a scientific breakthrough.
Incidentally -- Deming would also agreed with you on grades. Deming was
quite radical on grades. He believed that grades as we use them are
counterproductive to learning and development. Since grades do not help
us to improve but only measure performance post-facto, he felt grades
have no value. He believed that they reduce the quality of education by
reducing individual and group learning.
W. Edwards Deming would have agreed with everything you write here.
Warm wishes,
Ken
Ken Friedman, PhD, Disc (hc), FDRS
Professor
Dean
Swinburne Design
Swinburne University of Technology
Melbourne, Australia
--
Don Norman wrote:
The dangerous downside of Deming
(I just couldn't resist the alliteration)
Deming's practices and views have two sides. One is good and needs to be
even more widely applied.
The other is horrid -- and is of particular concern to two fields:
research and design. (Or any filed that requires delving into the
unknown or creativity).
Statistical Quality Control (SQC) is fine when you know precisely what
you are doing and when everything that needs to be controlled can be
controlled.
But I have seen devastating results when well-meaning managers tried to
apply these methods outside of their proper domains.
One manager of a research group where I once worked required the
research group to list the number of scientific breakthroughs they
expected to achieve each quarter, and then tried to hold them to it.
And the realm of engineering design tried to define the design process.
As one course syllabus I just read stated "grades will be based upon
adherence to formal methods. Intuitive design will not be permitted."
(Although at first I thought I agreed with this statement, it turns out
that my definition of proper process is precisely what they mean by
"intuitive" process. I agree that we don't want uninformed design, but
that's not what they mean -- they mean it has to fit the formal
requirements matrix and minimization methods.)
Design is complex. It has many stages, each of which might very well
require very different knowledge, skills, and techniques. Engineering
design, for example, algorithmic design, and SQC are completely
appropriate at some of the more technical components and during the
manufacturing stage. Much though I respect Denning's teachings and the work today that has
derived from them (we teach them in my MBA program), these methods are
the enemy when it comes to the place where designers or researches
normally flourish.
It is these managerial attitudes that make it so difficult for designers
to find their place in many companies, which is why so many major
products are so badly designed from aesthetic, suability, and even
functional grounds. And they are best kept away from research groups.
As for job shops and craft work: apply with great delicacy.
All tools have their opposite, conflicting sides. This is one example.
Don Norman
Breed Professor of Design, Northwestern University
Co-Director MMM Program. MBA + MEM: Operations+Design
Co-Director Segal Design Institute
[log in to unmask]
www.jnd.org
March and September-October 2009
Visiting Distinguished Professor
Department of Industrial Design
KAIST
Daejeon, Republic of Korea
|