Re: Postings and special cause(s)/Andrade

"Eduardo Andrade"

Tue, 8 Oct 1996 18:50:52 -0500

Would "out of statistical control" describe it?

Eduardo Andrade
eandrade@spin.com.mx

 =====================================================================

RE: Postings and special cause(s)/Marks

"Bruce Marks, DIS Ducom 868-3120"

Tue, 8 Oct 96 15:54:21 EDT

I did and X-bar and R chart on this data set with a histogram and a normal
probability plot.  I did not see any special causes, but the past history 
was limited.  In the lack of further past data, I attribute the variability
to common cause.

I'm only an Electrical Engineer, what do the statisticians say?

Bruce Marks
717/268-3120
marksba@ldcl01.dnet.dupont.com

=======================================================================

Re: Inventory control, large retail bookstore

dddsound@ix.netcom.com

Mon, 7 Oct 1996 21:20:43 -0700

[Moderator's Note:
Wil (dddsound@ix.netcom.com) asked a question through our Clemson CQI Server
that seemed more related to inventory control than Deming.  He responded
that is *was* the Deming approach he needed.  His response and original
message are shown below - he is not a subscriber, so please CC: him on any
replies.]

Dear Jim,

I would appreciate it if you would post the question anyway. I feel I need
direction from the Deming angle, considering the "stable process" I need to
maintain.

Best Regards,

Wil


ORIGINAL MESSAGE TO DEN:
>
> Subject:  Inventory control, large retail bookstore
> 
> TO: ALL MEMBERS
> 
> I have a client who is in desperate need to control inventory of his
bookstore.
> 
> The company is growing but is not turning inventory quickly.
> 
> The inventory is turning about 2 times. Average is 3 and good is 4
> in this industry.
> 
> 1) What elements should be studied to improve the "turns". How should the
program be set up?
>    What timespan should be used to measure?
> 
> 2) The system is stable right now although at undesirable levels
>    Is there anything that could be done right now to improve things
>    without affecting the stability?
> 
> Thank You For Your Help
> 
> Message posting through the Clemson CQI Web Server.

============================================================================

Re: Postings & Special Cause/Gogue

gogue@cri.ensmp.fr (GOGUE J.M. Societe MAST 39 50 99 67)

Wed, 9 Oct 96 12:31:20 +0200

On Tue, 8 Oct 1996 John David Kromkowski raised three important questions 
about Control Charts.

a. Central Limit Theory and Control Charts

> Following in tradition of jurists like Fermat,et al; I nonetheless propose
>following hypothesis: In a most direct sense CLT doesn't really have any
>thing to do with Control Charts, because sample size is generally too small
>(e.g., here the sample size is 2, not 9, i.e. range is of 2).  However, I
>propose, the CLT does interesect with Control Charts because one should (?)
>be somewhat skeptical of  control chart with only 9 values. 

Maybe there are several interpretations of the Central Limit Theorem. Long 
ago a professor at the university made me appreciate the formal beauty of the 
CLT. The demonstration aimed at explaining why different characteristics 
related to natural phenomena (e.g. size of individuals) were so often fitted 
with a single mathematical model : the Normal Distribution. I don't know 
any practical application of the CLT but the reasoning made me have a better 
understanding of how a system of causes produces variations (of the concept 
of common causes).

Deming stressed on the point that seeking to know whether the data are 
normally distributed is useless in the practice of Control Charts. Thus we  
should take it easy with the CLT. 

About skepticism  of  control chart with only 9 values, see below my (c).

b.  Control Charts Constants

>From where do the numbers, 2.66 and 3.267 (and 0 (D3?), come?  What
>is the THEORY behind these constants ("available from any reputable
>statistical handbook")? 

To my opinion, the THEORY is entirely based on the operational definition 
that Shewhart  gave of a stable system : the Ideal Bowl Experiment. The 
concept of three sigma limits comes from this definition. (See the book 
*Statistical Method from the Viewpoint of Quality Control*.) The constants 
are derived from the physical model of the Bowl through the Probability 
Theory.

c. Control charts with rare data

>The control chart only has nine values; this may be problematic. What's
>the theory and by what method?  Perhaps it would be a better practice to
>simple divide the first nine months in 27( 20?, 30?, 38?) equal parts and
>then provide the data?

I am not skeptical of  control chart with only 9 values. (I remember 
discussing this point with Don Wheeler who was not skeptical either.) 
But my problem is : what shall I do after my *judgement*?

The manager's problem is to make the best use of the available data to make 
predictions and decisions. With only 3 values, you may have the intuition 
that the characteristic is stable or not, but it's just an intuition of 
course. The more you have values, the more your intuition moves toward a 
rational prediction. When I *judge* that the system is out of control, 
it means that I have to look for a special cause. Thus I should decide 
to make an investigation taking time and money. In many cases, I don't 
take such a decision because the confidence level is too low.

My belief is that the * individual values and moving range* control charts 
are the most effective charts when data are rare (e.g. one per day, week or 
month). The control limits must  be often re-calculated with the new data.
The book *Understanding Variation - The Key to Managing Chaos* by Don 
Wheeler is very clear in this respect.

Finally, I fully agree with the JDK calculations.

Jean-Marie Gogue
Versailles France
gogue@ensmp.fr

==============================================================================

Re: Systemic -vs- systematic/Woods

John Woods

Wed, 09 Oct 1996 08:14:37 -0500

Steve Delph wrote:  
>I believe one of the stumbling blocks we may have along the way is this idea
>of thinking of our agency as a "system" and how the agency should function
>and be managed as a "system."

I'd say, Steve, that coming to that realization is the first step toward
transformation, but it has to be appreciated by top management as well.  The
fact is that your agency has always been a system, but if you stick to
traditional management approaches, attitudes, and techniques, you will
compromise the performance of the system.  There will be more mistakes made,
more wasted resources, less satisfied customers of your services than there
could be.  Some people call the Deming approach "quality management."
Sometimes I like to call it "smart management."  It's just what you do when
you begin to appreciate that your agency, or any organization, is a system.

John Woods
jwoods@execpc.com

===========================================================================

Postings and special cause(s)/Mossman

am249@cam.ac.uk (Alan Mossman)

Wed, 9 Oct 1996 09:10:21 -0200

Jim

Is your message a special cause ?  Is October "falling behind" ? ;-)


Alan

[That which is measured, improves - eh, Alan??]
--

                                Alan Mossman
_______________________________________________________________________

Cambridge Programmes for Industry       | weekends and holidays
1 Trumpington Street Cambridge CB2 1AQ  | 19 Whitehall Stroud GL5 1HA
internet:am249@cam.ac.uk                | 100733.3202@compuserve.com
voice: (+44) 01223 332722               | voice   01453 765611
fax: (+44) 01223 301122                 | fax     01453 752261
evenings: 01223 322162                  |
________________________________________|______________________________

======================================================================

Re: Control Chart Usage/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

09 Oct 96 09:19:26 EDT

Prevette:  In a situation, such as you describe, where you are
a kind of prophet crying in the wilderness,  it is very important
that you develop some kind of historical record.  The best thing
is a run chart with such simple entries as:  Number of control
charts produced per week for the last year.  Number of requests.
Number of control charts which produced an action.  Number of
requests for training in SPC.   A run chart, kept up to date
on display in your office will help you to see the progress
you have made.

In Japan I found charts like these a number of places.   They
represent the application of quality methods to the process of
introducing quality.

In short,  when you teach about quality,  do it with quality.

Myron Tribus,  350 Britto Terrace,  Fremont, CA 94539
Ph:510 651 3641  Fax: 510 656 9875
I find that when I soak up information like a blotter
I often get it backwards.

===================================================================

Re: Systemic -vs- systematic/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

09 Oct 96 10:16:37 EDT

You can develop ways of "thinking of the enterprise as a system" by 
producing deployment flow charts of the most significant processes 
used by the enterprise.   The interconnectedness,   the interdependencies,
the need for treating internal customers as important,  etc., will
almost leap off the page.  
If you do not know about Deployment Flow Charting,  drop me a note 
off-line.

Myron Tribus

=======================================================================

Re: Postings and special cause(s)/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

09 Oct 96 10:06:30 EDT

It satisfies the intellect to know more about the theory underlying the
control chart.  When I first learned about SPC it was during WWII and I
had no interest in theory.   Then in 1961,  after publishing my book
showing how to derive thermodynamics from information theory,  I
was attacked so much by people who thought thermodynamics was a branch
of religion, I was forced to study probability theory intensely.  This
led me to examine the connction between entropy and the central limit
theorem.  The connection,  mathematically supported,  is in my second
book (Rational Descriptions, Decisions and Designs,  Pergamon, 1969)
which is now out of print but perhaps available in the part of your library 
where people seldom go.

When I relearned SPC after meeting Dr. Deming,  I also had access to
computers and,  after reading Shewhart's description of what he had done
said to myself,  "Aha!  Now that we have computers and  the maximum entropy
estimate (which agrees with the Central Limit Theorem) we can dispense
with these Shewhart numbers which are so handy for rapidly finding the
upper and lower control limits,  etc.  After all,  Shewhart produced them
because in his day finding the sum of squares and all that was a terrible
chore and certainly beyond the people on the shop floor."  (I often talk
to myself in long sentences)

So I experimented with tables of random numbers,  creating data and then
applying SPC techniques to their analysis.  I was chagrined to find that
no matter what I did,  it was faster and easier to use those little numbers
that Shewhart had generated than to do it the "correct" way.   
The differences between his "approximate" method and the "correct"
(i.e.,  CLT or Max Entropy) was less than the width of a pencil line 
(I always end up making graphs.  It is better for the brain to 'see' the data).

So,  I quietly shelved what I had written and decided my time would be better 
spent on something else.

Where did the Shewhart numbers come from?  Experiments,  I guess.
 I know of no theory.  Does anyone else?

For those who are deeply interested in this subject,  I can suggest nothing
better than study of Don Wheeler's books.  For the connection among
CLT, Entropy estimation,  Thermodynamics,  uncertainty,  and the theory of
information,  masochists may wish to read my book.

  The original bowl and chips Shewhart used are now
on display at the American Statistical Association Offices.  

BTW,  the American statistical Association is supposed to have a replica 
of Dr. Deming's office from Washington DC.  This was planned some time ago 
but I have not been to their HQ to see for myself.  .

[Moderator's Note:  Yes - The ASA *has* a replica of Dr. Deming's office
in their DC office area - at previous Institute Conferences, tours were
scheduled.]
 
Myron Tribus

=====================================================================

Re: W. Edward's Deming Medal Nominations/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

09 Oct 96 10:06:27 EDT

Dr. Deming was very proud of the existence of the Deming Prize in Japan.
He even donated his royalties from sale of the translation of his lectures as 
a way to get the prize started.

The prize is NOT a competition.   There may be more than one winner.  

I do not know if there is to be more than one Deming Medal each year.

Myron Tribus,  350 Britto Terrace,  Fremont, CA 94539
Ph:510 651 3641  Fax: 510 656 9875
I find that when I soak up information like a blotter
I often get it backwards.

======================================================================

Re: W. Edward's Deming Medal Nominations/Latzko

"William J. Latzko"

Wed, 9 Oct 1996 14:00:51 +0000

At 12:18 10/8/96 +0000, you wrote:
>
>Do I understand Deming less than I thought or wasn't he against awards.
>
>Jack Campbell
>Cycle Time Reduction Facilitator
>

Dr. Deming was against awards as an incentive to get people to work. He was
the first recipient of this medal and personally handed it out to the other
recepient whenever he could. While he was alive, he was an active member of
the selection committee.

Bill Latzko
---------------------------------------------------------------------------
William J. Latzko
215 - 79th Street
N. Bergen, NJ 07047
Voice:     201-868-5338
Facsimile: 201-868-5338
E-mail:    latzko@worldnet.att.net
------------------------------------------------------------------------------

===========================================================================

Re: Postings and special cause(s)/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

09 Oct 96 09:19:28 EDT

It's hard to be sure graphics reproduces in e-mail.  But,  measure time
downward from top of the figure.   Horizontal axis is number of postings.


     |       *
     |      *
     |               *
     |              *
     |            *
     |                 *
     |            *
     |      *
     |                          *
     _____________|_________________
            ^            ^     

The "^" mark + and - 1 sigma.  Hard to tell if the process is under
control, only 9 data points.   Cannot claim that the one point at
the end means anything.

If you do not print this in monospace type (courier or monaco) the graph
means even less.

Myron
===============================================================================

Re: DEN Challenge/Taylor

Morris Taylor

Wed, 9 Oct 1996 23:15:15 +0100

I enjoyed reading the post from David Snook-Luther and all the other
contributions to the thread. 

I am disappointed that I will be unable to join the discussion this weekend.  

I do however have several suggestions that I hope will prove useful in some way.

If we were to develop a (very short) set of ground rules which tightly
addressed the topic of managing the subjects and titles of threads then the
'deep den' would automatically come into existence within the present
framework. If subscribers had self discipline and self management with
regard to their titling and if Jim The Moderator were to either return to
sender or retitle contributions which 'almost' matched the thread of the
subject, we would not require to have a separate list at all. 

Most of us will already use the database sort function on most mail handling
packages.

I have found it interesting to watch the way in which we often diverge away
from the topic by giving a satisfactorily titled thread a new name! This
makes it difficult to sort on the mail handler as well as difficult to
follow the thread of the arguments as they unfold.

My suggestion is therefore very simple: 

- Stick to the existing title unless you have a new thread.

- If you have a new thread, suggest a title for your new submission to the
moderator. Leave it to him to decide the title of your masterpiece. He may
want to change the title for reasons which make more sense to him than they
do to you. Allow him that discretion. Try to save him the trouble by being
intelligent when you make your submission.

- Ask the moderator for a title.

- Try to make a contribution to the long suffering Jean-Marie who diligently
puts in his weekly request which most of us (including myself) only make the
effort to respond to if it looks a 'trendy' topic. 

If anything is worth doing it is usually damn hard work. This Deming stuff
is no exception.

Have a good weekend.

What will you do Monday?


Sincere Regards
 	
Morris Taylor 

Aberdeen Hypnotherapy Centre
at
MOTIVATION TECHNOLOGY
One Belgrave Terrace	
ABERDEEN				
UK
AB25 2NR

+44 (0)1224 633 222 	(Voice)
+44 (0)1224 646 999 	(Fax)

==========================================================================

Re: Postings and special cause(s)/Kilkenny

Matt Kilkenny

Wed, 9 Oct 1996 11:35:00 -0700

Read Wheeler's books, as he clearly explains how Shewhart "derived" how
the control chart works.
Things to think about concerning the control chart and the central limit
theorem:

How could an individual moving range chart work with non-normal
distributions? As Wheeler explains, Triangular shaped distributions
still have 99% of their distribution within +-3 sigma (not purely
probability limits, but limits that are based off of empirical evidence
that they work with a foundation in probability) which is the economic
balance from making type 1 and type 2 errors.
I asked my professor in my SPC class this question and he sidestepped
the issue.

Wheeler also shows how the central limit theorem doesn't hold for a
range chart as (I think) the subgroup size increases.  So if it doesn't
work, then how could it work because of it.

Another interesting note, probably the most popular "SPC" book by grant
& levenworth (newest edition) spends chapter 3 "Why the control chart
works" on topics about taking samples from any shaped distribution
including triangular, that eventually the sample distribution will be
bell shaped.  It also talks about the normal curve, central limit
theorem etc.  I don't remember reading a statement in this chapter here
that "the normal curve, central limit theorem is why the control chart
works."  Most people will assume this is why it works based on the title
of this chapter.   This is interesting because in chapter 5
"Fundamentals of probability" it has a section on central limit theorem
which says, "while this fact is very useful in analyzing probabilities,
it does not form the basis for Shewhart control charts with +-3 sigma
limits" and it refers to Wheeler's material.  I told Dr. Wheeler this
and he was surprised to hear this acknowledgement in grant &
levenworth's book.

On the one hand  the book seems to imply the control chart is based on
the central limit theorem, but on the other hand there is a little blurb
in another chapter that says it isn't based on the central limit
theorem, it seems to "weasel".  Maybe this a grudging acknowledgement by
Grant & Leavenworth that it was a mistake through all their old editions
(many people taught) that it's wrong to state that the control chart
works because of the central limit theorem.

This is rule #4 of the funnel experiment, the blind (readers of grant &
leavenworth) leading the blind (readers teaching people or new authors
on control charts who copy without researching.)


Matt Kilkenny
matt.kilkenny@readrite.com

=============================================================================

Re: Systemic -vs- systematic/Constantine

John Constantine

Wed, 09 Oct 1996 08:34:04 -0700

Something does not jive here.

The agency has just recently adapted TQM as its management 
philosophy?...and you are being told you are carpet dirt????

Seems like an age-old problem of bureaucracy...lack of communication 
leading to imposition of rigid controls on change. Yours is a fear-based 
system, I'm afraid, and one which appears to be bent on control and not 
improvement. God help you.
-- 

Regards, 
John Constantine
Rainbird Management Consulting
PO Box 23554
Santa Fe, NM 87502
http://www.trail.com/~rainbird

==========================================================================

messianic murmurings/Archambault

Blaine Archambault

Wed, 09 Oct 1996 13:03:37 -0500

I have been "lurking" on this list for about two months now.  I am also disturbed by 
"messianic murmurings" from anybody.  I am very involved with a new political party in 
Canada, The Reform Party of Canada.  Through this experience I have learned that the 
soft sell works better than the hard sell.  I also pass along relevant ideas that come 
out of this list to the QA community in my company.  I don't pass along zealot 
comments.

Blaine Archambault
President
Ottawa Centre Riding Assoc.
Reform Party of Canada

blaine@netcom.ca
blaine@newbridge.com

=========================================================================

Re: Systemic -vs- systematic threads/Tolman

"Anton O. Tolman, Ph.D."

Wed, 9 Oct 1996 10:49:06 MST

In a recent posting regarding the "Deep DEN" idea, JD responded to 
Julie by saying:

> Systemic v. Systematic
> 
> Three cheers for holism and SYSTEM, but at some point its important to do and
> think about the funnel, the red beads, the theory behind control charts and
> their use, the philosphy of knowledge, the essence of what it means to be a
> learning organization or person, and about the specifics of the idea of
> System.
> 
> Note also that my proposal was to alternate between free form and focus.
> Hence, I am not altogether certain that any of your objections (new comers,
> tuning out, "richnes of // dialogues") are well founded.  Of course, I may be
> wrong but all of the things you mentioned may be the outcome of what we have
> now.
> 
> Finally,  "richness of the parallel dialogues".  You know, counterpoint is
> rich, polytonality is rich, polyrythm is rich.  All require some rigour.
>  Making the following a metaphor for the DEN -- while either Plutarch or
> Thelonious Monk said to understand harmony one must investigate discord,
> whether or not stirred vegetable soup is rich depends on whats in it.


I think that this is an interesting discussion, and I have been 
reading trying to decide what my opinion is.  At this point, as a 
"beginner" I am very interested in soaking up as much as I can, yet I 
think I would benefit greatly from more in-depth discussion of 
various parts of the SPK; sometimes, to begin learning, I need to 
break things down, need to define things clearly in my mind.  I have 
found this out as I have begun training in some of the principles to 
large groups of managers in my organization.  The clarity and 
definitions needed to teach have helped me to better understand.  At 
the same time, if we are to value and learn anything from WED and 
other systems thinkers, we must apply systems thinking, which humans 
are not that good at (yet).  This is right-brain stuff, which I 
sometimes struggle with.

I thought of throwing out the following considerations:

*  establish a thread for a deeper, more focused discussion.  The 
title of the thread would remain constant (constancy of purpose? ) 
so that all who saw it would realize that it was a focused 
discussion.  Perhaps even the title of the thread could relate this, 
e.g. Focus: 14 points.  This thread could be embedded in the flow of 
the DEN normal messaging process.  That way, if some readers found it 
too focused or too confusing, they could just delete those messages 
and focus on the others.  For other readers, they could participate 
in both arenas and so benefit.  
*  Why don't we utilize the PDSA cycle ourselves for this?  Consider 
our current discussion part of the Plan.  Reach a consensus on the 
intervention, implement it for a trial period, say 2-3 months, and 
then study the outcome.  If it seemed to be a benefit, continue.  If 
not, discontinue.

I'll be interesting in hearing your responses.

________________________________
Anton O. Tolman, Ph.D.
Director, Psychological Services
Director, Quality Management
Wyoming State Hospital
P.O. Box 177
Evanston, WY  82931-0177
Anton@wsh.state.wy.us
(307) 789-3464
---------------    

================================================================================

Lessons from Religion/Beare

dbeare@zetnet.co.uk (David Beare)

Thu, 10 Oct 1996 05:28:31 +0100

As a newcomer to Deming,I am very much attracted to the ideas although my
understanding is still peripheral. I have been reading the Denlist for a
month or so now and was pondering over the characteristics of the
communciation. The impression I have is:

A devoted following of people who admit to not fully understanding the
meaning and impact of Deming but are dtermined to continue the search. Great
difficulty in expaining it in currently understood worldly terms. The need
for a leap of faith and commitment to get the full benefits of the
philosophy, possibly at the expense of wordly things such as career,
popularity, and respect.

It seems to me that this is similar to the characterstics and problems of
religion! Maybe the various comments about a sort of messianic devotion to
Deming (sometimes at the expense of other learning) are not so far out. If
so, is there something for us to learn from religion about spreading the
message; long timescales (transcending an individual's lifespan), faith,
sacrifice, etc.)?

Just a creative thought. Anyone care to take up the theme?

Regards,
                 David

Email: dbeare@zetnet.co.uk
Tel:  01224 321550 (evening/weekend)
       01224 883092 (day)

==========================================================================

Use of Surveys/ Prevette

Steven_S_Prevette@RL.gov

Thu, 10 Oct 1996 07:52:48 -0700

     I have been having an off-line discussion with Geoff Penrose about the 
     use of personnel surveys -> measuring worker attitude -> determining 
     common causes & special causes (as appropriate to the statistical 
     analysis of responses) in order to improve worker attitude, and 
     therefore business success.  I will admit that I have been trained in 
     survey construction and statistical analysis of survey results.  
     However, I have not (yet?) been involved in an application of survey 
     results (either of workers or customers) that seemed to actually cause 
     a significant improvement in business success.
     
     I have supposed the following:
     
     Dr. Deming stated that 94% percent of problems a business faces are in 
     the processes and systems (which are owned by management) and 6% in 
     the workers.  Perhaps if I improve worker attitude, then I can improve 
     on the 6% part.  But working only on worker attitude will not improve 
     the remaining 94% directly.
     
     Attitude of the workforce is still important.  Dr. Deming assessed the 
     attitude of the workers by group interviews/interactions and by 
     looking at sick day records.  To my knowledge, Dr. Deming did not make 
     use of formal surveys.  "Poor" attitude on the part of the workers was 
     said to be caused by poor systems and processes.  Fix the systems and 
     processes (which are owned by management, but will require input by 
     the workers), and worker attitude will improve.
     
     I will admit that Dr. Deming's theories address "driving out fear", 
     and determining what the fears are may be assisted by survey.  
     Overall, surveys may help to identify a problem, but I think the fix 
     is more through directly working on the processes and systems rather 
     than directly focusing on attitude.
     
     The question to DEN members is:  does anyone have experience with the 
     use of personnel survey results as part of an application of Deming 
     management theory, and more importantly, an application that lead to a 
     significant improvement in business success?
     
     - Steve Prevette
       steven_s_prevette@rl.gov

=======================================================================

Re: Customer expectations/Carlson

"fury1"

Thu, 10 Oct 1996 09:23:25 -0500

Dr. Tolman replied to Julie Beedon's post which mentioned her comment about
Quality being what all the stakeholders say it is.  Dr. Tolman argued
against that by using the example of a hospital with no customers but which
made all the other stakeholders happy.  This is an example of reducing an
argument to an absurd level, demolishing the absurdity, and then stating
you have demolished the original point.  (This technique is frequently used
in political campaigns, by the way.)  Julie's point is still valid, since
no organization without customers can satisfy the remaining stakeholders.  

Quality is an ambiguous term at best.  It means so many different things
and can be defined so many different ways as to be virtually meaningless. 
Deming himself defined quality both in terms of the customer and in terms
of satisfied employees.  Both are valid, yet both do not have to be true at
the same time.  A Swiss watchmaker may have had great pride in his work,
but the customer demanded greater accuracy at lower cost.  Digital watches
made proud Swiss watchmakers obsolete.  

We ought not to stab each other over definitions of quality.  Like
pornography, I know it when I see it.  

Kris Carlson
fury1@wave.beaches.net

==================================================================

Re: Lessons from Religion/Woods

John Woods

Thu, 10 Oct 1996 07:28:24 -0500

David Bear asks:
>Is there something for us to learn from religion about spreading the
>message; long timescales (transcending an individual's lifespan), faith,
>sacrifice, etc.) [with regard to spreading the insights of Deming regarding
>management]?

Here's what I would say about that:  First of all, we need to appreciate
that religion usually has more to do with ideology and tradition than with
doing things better.  The tenets of most religions are not up for question.
Traditional management practices, in this way, are similar to a religion.
People take them on faith and never think about whether there might be a
better way of doing things.  

This is quite the opposite of what Deming was talking about, which is more
appropriately compared with notions of spirituality.  Spirituality is about
asking questions about what animates us as human beings.  It helps us
appreciate that we are not separate from the world but a part of it.  If
fosters openness and exploration of our human nature.  It reminds us that by
looking out for others in a compassionate and open-hearted way, we look out
for ourselves at the same time.  

In a manner of speaking, quality management is about similar issues in the
context of an organization.  To the degree that TQM has validity, it is
because it helps us manage our human nature effectively in organizations.
It provides a method for understanding how well we are doing at working
together (and we are working together--in an interdependent manner, just as
we live together on the planet--interdependently), and a methodology for
improving.  It reminds us that organizations prosper or wither depending on
their ability to add value for segments of society (customers) and society
as a whole.  

Will people eventually come to understand the validity of the ideas
articulated by Deming and many others?  You can count on it.  Quickly?  I'm
not sure of that.  Why would I say this?  Because this idea of Deming is a
better articulation of how to work effectively in the world.  It facilitates
survival and growth rather than demise.  Any management techniques that are
not aimed at making the organizational system operate ever more efficiently
and effectively to serve customers and society as a whole will eventually
die out.  

One thing you can be sure of: the appeal of Deming's ideas has absolutely
nothing to do with Deming personally.  His are useful articulations of
principles that have always been at work when people organize themselves to
achieve some goal.  If you start from the perspective of the organization as
a system (that is, a perspective that reminds us that we are all in this
together) and then work out the implications of that, you will eventually
come up with the tools and techniques of TQM.  These ideas are not true
because Deming said they were.  They are true because they are true.  

So let's not compare TQM with religion.  That's a bad precedent.  Too many
religions are about corrupting the truth for personal gain and power in
exactly the same way as political ideologies. Such ideologies are formulas
for failure, not to mention heartache, fear, misery and other human
self-induced problems.  However, we can compare TQM to the search for one's
spiritual nature, but in doing so, we must appreciate that that search
mainly is about understanding of our human nature as it manifests itself in
our day-to-day lives.  And, again, TQM is about using insights in our human
nature to perform ever more effectively in organizations.  

John Woods
jwoods@execpc.com

=======================================================================

Reply: Postings and special causes from Kromkowski/Look

"Look, Alson"

Thu, 10 Oct 1996 08:58:00 -0400

Comments.

1.  The first moving range should be 24.
2.  Are the data in time order?
3.  What is the nature of the data?  Have we rationally sampled the
data?  If the data are not rationally sampled then it may not make
sense.  It is important to sit down and think about the process,
sources of variation,  and so forth.
4.  UCL mr=203 (approx),  UCLx=294,  LCLx=-36.45.  Xbar=128.8,
Mrbar=62.13.
5.  If the data are rationally sampled the last data point on the moving
range chart (203) is statistically out of control.
6.  Statistical theory was used to guide Shewhart (Tchebycheff's
theorem, law of large numbers, experience,......) were used to develop
what we today call Shewhart control charts.  Shewhart (1931 p 277)
7.  Shewhart sort of hints at the CLT (1931 p 316 I believe) but points
out that in the real world you don't really have normal (gaussian)
distributions.  Grant Leavenworth in their famous book on SQC point out
that control charts work due to experience that you will find special
causes more often than not with 3 sigma limits than 2.
8.  The CLT becomes important if you want to use some of the other
statistical out of control rules such as 2 out of 3 outside of 2 sigma
and so forth.
9.  Shewhart indicates on p315 (1931) that you can use his method for
subgroups (rational sampling) as small as four and as few as two to
detect a lack of statistical control.  So you can use his method with
very few samples of data.  Obviously the more data you have the better
you can judge statistical control.  It is a question of knowledge of the
process and the data analysis.
10.  From what I understand Shewhart in the 1920s and 1930s developed
the p chart and then the xbar and s chart where the s was what we today
call the root mean square deviation.  This of course was difficult to
implement.  Sometime during the 1930s a British statistician by the name
of Tippett (I think thats how you spell his name)  showed how to use the
range to estimate the standard deviation.  Around that time and it is
unclear to me an American statistician by the name of Paul Hoel showed
that you could use moving ranges to estimate the standard deviation.
Some of the history maybe lost or  it maybe that I have been
unsuccessful in piecing together the history.  During the time frame of
WWII  courses were taught by Deming, and Grant.  In 1946 after the war
Grant published his now famous book on SQC.  In this book it contains
xbar and r as well as other control charts.  Soon after Bonnie Small at
ATT published the Western Electric book.  Both books have the xbar and r
and x charts.  The 2.66, 3.27, and so forth are derived in Irving Burr's
SQC book and to some extend in Grant Leavenworth.  They are based on
normal (gaussian) theory.  The derivation can also be found in the ASTM
STP 15 booklet (I believe).  Does the fact that we are using possibly
non-normal data with formulas that have some basis in normal (gaussian)
theory detract from their use?  No.  As John Tukey and many others have
pointed out there are no gaussian distributions in the real world.
Shewhart's control charts work because they have some theory behind them
and experimental verification.
11.  Don Wheeler's books are a good place to start.....

Alson C H Look
EJLA29B.prodigy.com
Standard disclaimers

======================================================================

Re: Authorized Thinkers and back-sliding/McAra

David_McAra@scet.org.uk (David McAra)

Thu, 10 Oct 96 08:00:45 -0000

D> Good News/Bad News
D> The good news is that I cannot go back.  The bad news is also that
D> I cannot go back, which makes it very hard sometimes to operate in
D> today's companies or society.
D> 
D> Mike Newman Downstrm@aol.com

The other day I noticed some of the words in Bob Marley's "Redemption Song."

"Though it feels like a losing battle, there can be no retreat.  Won't you
help to sing these songs of freedom?"  

David McAra, Effectiveness Consultants UK



--

     _/_/_/  _/_/_/  _/_/_/  _/_/_/  | Tel: (0141) 337 5000  
    _/      _/      _/        _/     | Fax: (0141) 337 5050  
   _/_/_/  _/      _/_/      _/      | Net: scet@scet.org.uk 
      _/  _/      _/        _/       | AppleLink: SCET.DEV   
 _/_/_/  _/_/_/  _/_/_/    _/        | WWW: http://www.scet.org.uk
......learning through technology   |

========================================================================

Maastricht/Gogue

gogue@cri.ensmp.fr (GOGUE J.M. Societe MAST 39 50 99 67)

Thu, 10 Oct 96 17:54:05 +0200

On Tuesday, 8 October 1996, William Vickrey, Nobel Prize of Economy 1996, 
said that the Maastricht rules for the European currency are a piece of 
nonsense. (Press conference.)

I agree. These rules are pure Management by Objectives. They set numerical 
goals for January 1999. They don't care about the process.

But the aim of a European Federation with a single currency is respectable.

Jean-Marie Gogue
Versailles France
gogue@ensmp.fr

====================================================================

Re: Lessons from Religion/Holt

Steve Holt

Thu, 10 Oct 1996 08:53:51 -0700

Yes, there are things to be learned from studying religious movements.
Social and cultural change are fairly universal.  For an interesting
example, albeit in a different field, take a look at Guy Kawasaki's book,
"Selling the Dream : How to Promote Your Product, Company, or Ideas-And Make
a Difference-Using Everyday Evangelism."  

Guy was one of the Macintosh development team responsible for finding
software developers. He realized early on that he needed "evangelists" to
push the computer and software. He actually signed up for a class series at
a fundamentalist seminary (if I remember correctly)  to study Christian
evangelism so that he could convert their methods to the computer world. 

In fact, he's still at it. To see the latest incarnation, check out his web
site at:
http://www.evangelist.macaddict.com/

Steve Holt
steven.c.holt@boeing.com

======================================================================

Re: W. Edward's Deming Medal Nominations/Kilkenny

Matt Kilkenny

Thu, 10 Oct 1996 08:49:00 -0700

I heard he donated the money to translate some statistics books like
Shewhart's into Japanese.  I don't remember reading about him directly
donating his money to the Deming prize.  By the way I'm trying to write
a paper for school on how they calibrate Malcolm Baldridge Award
examiners/judges?  ( I don't think they even do)  Does anyone have any
suggestions? I would also be interested on how Deming Prize examiners
are calibrated.

Matt Kilkenny
matt.kilkenny@readrite.com

[Baldrige and most state quality processes employ a calibration process
in the consensus-based scoring process.     Jim Clauson]
 
============================================================================

Re: Lessons from Religion/Clark

"Don Clarke"

Thu, 10 Oct 1996 09:11:34 -0600

Accepting for just a moment the comparison of attributes between 1) people
who seek to understand Deming's teachings with 2) those who are religious,
we should remember that "if it talks like a duck, walks like a duck, and
looks like a duck--it is not necessarily a duck".  Halloween nears.

A few other comments.  Because I don't understand fully all of Frege's
logic does not mean that I can't use to my benefit tools of his logic that
I do understand.  Similarly, although I may not know all of what Deming
meant by SPK, I find his 14 points, elementary SPC, funnel experiments,
elements of psychology, etc., understandable and beneficial.  Thus I find
no leap of faith necessary to accept Deming's work, and his SPK would
reject such a foundationless leap.  Also, it's easy to be popular, and I'd
hate to have a career in which I wouldn't apply knowledge to add value from
tools that work.  That wouldn't be self-respecting.  

Finally, I don't consider *messiah* a detrimental term.  It's a term that
can be used for a leader of a hoped for outcome.  I see no problem in
accepting Deming as a symbol for constant progress towards a better way of
managing and working and living.  We revere many other personalities for
lesser reasons.

=================================================================

Re: W. Edward's Deming Medal Nominations/Latzko

"William J. Latzko"

Thu, 10 Oct 1996 14:56:15 +0000

At 14:06 10/9/96 +0000, you wrote:

>I do not know if there is to be more than one Deming Medal each year.
>
An important point. There is provision for more than one winner.

Bill Latzko
---------------------------------------------------------------------------
William J. Latzko
215 - 79th Street
N. Bergen, NJ 07047
Voice:     201-868-5338
Facsimile: 201-868-5338
E-mail:    latzko@worldnet.att.net
=====================================================================

Re: Postings and special cause(s)/Latzko

"William J. Latzko"

Thu, 10 Oct 1996 14:56:11 +0000

At 14:06 10/9/96 +0000, you wrote

>Where did the Shewhart numbers come from?  Experiments,  I guess.
> I know of no theory.  Does anyone else?

The factors came from sample theory. An explanation of the method to compute
these factors can be found in Supplement A and B of MANUAL ON PRESENTATION
OF CONTROL CHART ANALYSIS ASTM Manual Series MNL 7 published by the American
Society for Testing and Materials. The basis for the various factors are
mentioned by Shewhart in his book, THE ECONOMIC CONTROL OF QUALITY OF
MANUFACTURED PRODUCT, p. 184. Shewhart based his work on that of Student (W.
S. Gosset), Pearson and Fisher. For the factors of the range, Tippet and
Pearson worked on this aspect. Their results were published in BIOMETRIKA
TABLES FOR STATISTICIANS Vol I by Pearson and Hartley (see page 42 ff for a
discussion of these tables.) Further theory can be found in Kendall and
Stuart's, THE ADVANCED THEORY OF STATISTICS page 338 ff.

>
>For those who are deeply interested in this subject,  I can suggest nothing
>better than study of Don Wheeler's books.  For the connection among
>CLT, Entropy estimation,  Thermodynamics,  uncertainty,  and the theory of
>information,  masochists may wish to read my book.

While Don Wheeler's book gives a great start, my feeling is that to study
the original work of Shewhart cited above will answer more questions for the
serious student. The concept of what the control chart is meant to do and
what it is not meant to do is explained by Shewhart. Personally I like to go
to the source to avoid funnel rule 4.

>
>  The original bowl and chips Shewhart used are now
>on display at the American Statistical Association Offices.  

I believe that the bowl was donated by Professor Ott's wife and resides at
The American Society for Quality Control in Milwaukee. I saw some of the
coat tags that Shewhart used on display with the bowl.

>
>BTW,  the American Statistical Association is supposed to have a replica 
>of Dr. Deming's office from Washington DC.  This was planned some time ago 
>but I have not been to their HQ to see for myself.  .
>
>[Moderator's Note:  Yes - The ASA *has* a replica of Dr. Deming's office
>in their DC office area - at previous Institute Conferences, tours were
>scheduled.]
> 
The ASA has a replica of Ed's office at their headquarters. Many of the
books that he used are there. One can apply to the ASA to use the office for
research. I do not know all of the restrictions but they can be obtained
from ASA.

Bill Latzko
---------------------------------------------------------------------------
William J. Latzko
215 - 79th Street
N. Bergen, NJ 07047
Voice:     201-868-5338
Facsimile: 201-868-5338
E-mail:    latzko@worldnet.att.net

========================================================================

Re: Deployment Flowcharting/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

Thu, 10 Oct 96 23:26:38 EDT

Quite a few people  (well, two) have asked me to explain deployment flow
charting.    The idea is very simple.  There are software programs designed
to take all the work out of it.  You will find a very, very brief description
of a deployment flow chart in Peter Scholtes TEAM HANDBOOK on page 2-24.

The videotapes,  "Deployment Flow Charting" (Vols I and II) distributed by
Quality and Productivity,  Inc.  (310) 824 9623 are devoted to examples of how
to use them,  how to construct them,  how to analyze processes,  etc.

There are quite a few examples in the collected essays,  "Quality First"
available from SPC Press or from the National Society of Professional Engineers,
1420 King Stree, Alexandria, VA 22314   (703) 684 2800

Myron Tribus

===================================================================

Re: Postings and special cause(s)/Tveite

MDTveite@aol.com

Thu, 10 Oct 1996 18:25:17 -0400

There have been a number of messages recently which connected the Central
Limit Theorem and the genesis of control charts.

I was once with Dr. Deming when someone tried to make such a connection.
Deming asked, "what is the Central Limit Theorem?" His questioner quoted it
to him, whereupon Deming commented that such a theorem was of a different
world. The central limit theorem applies to the average of a number of iid
(independent, identically distributed) random variables, which is the
assumption (almost universally made in traditional statistics) that Shewhart
was trying to test:  Are, in fact, the observations independent and from the
same distribution? Shewhart certainly would not have assumed that which he
was testing!

My studies of Shewhart have shown that he settled on his control limits as,
*in his experience*, the limits which minimize the two mistakes (treating
common causes as though they were special, and treating special causes as
though they were common). Also, as far as I can tell, he never made
assertions about the form of any distributions for observations from
processes which haven't been brought into a state of statistical control.

Mike Tveite
mdtveite@aol.com

======================================================================

Re: Postings and special cause(s)/Tolman

"Anton O. Tolman, Ph.D."

Thu, 10 Oct 1996 14:52:12 MST

I have enjoyed reading some of the responses to Jim's post about the 
number of messages on the DEN.  No wonder I had so much trouble 
keeping up with reading in September!  

I think that there may have been a typo in the first range calculated 
for the XmR chart in that it showed a range of 4 from 86 to 62 
instead of 24.  I am not sure this would have made a significant 
difference in the control chart or the contro limits.  It appeared to 
me that the system was stable, but looking at Myron's "vertical" 
control chart, I was struck by the fact that we will only know the 
full story with several more months' worth of data.  

If October's data shows regression to the mean and falls back, then 
the high September number is random, common cause variation.  If 
however, October's data is also close to the 3 sigma + side, and so 
is November's, then we could begin to consider Wheeler's rules for 
determining if a system is out of control even if there are no out of 
control points.  This could indicate a significant shift in the 
process and it might be intriguing to determine what that shift was, 
since I don't believe we entered a PDSA cycle and are purposely 
trying to shift the mean.  

So, I would encourage Jim to continue posting for at least the next 
few months the monthly totals for us to compare.  I look forward to 
it.  

[The DEN is custmer focused - so you *will* get the data.  Jim Clauson]
________________________________
Anton O. Tolman, Ph.D.
Director, Psychological Services
Director, Quality Management
Wyoming State Hospital
P.O. Box 177
Evanston, WY  82931-0177
Anton@wsh.state.wy.us
(307) 789-3464
---------------    
====================================================================

Re: Lessons from Religion/Harris

egharris@bconnex.net (Emily & Grant Harris)

Thu, 10 Oct 1996 16:12:30 -0400 (EDT)

 In his message David Beare asked:

>... is there something for us to learn from religion about >spreading the
message; long timescales (transcending an >individual's lifespan),
faith,sacrifice, etc.)?

Interesting thought and makes a connection with a little book that I
constantly recommend. It's "Scuttle Your Ships Before Advancing (and Other
Lessons Learned from History on Leadership and Change for Today's Managers)"
by Richard Luecke (Oxford University Press, 1994). Chapter 4 is entitled
"The Power of Ideas: Two Reformers Change the World". One is Martin Luther
and the other is Dr. Deming. 

The final sentences from this chapter are worth noting, especially in the
context of "messianic Deming followers". 

"They [people with ideas]  are the people who make us angry and
uncomfortable with their methods, and our instinct is to ignore or get rid
of them." 

I have often said that I feel like an early Christian. I'd just prefer not
to be eaten by the lions!

Emily Harris
egharris@bconnex.net
 
===========================================================================

Lessons from Religion/Harris

Thomas Harris

10 Oct 96 11:12:54 CDT

From: Tom Harris
David's comparison of Deming followers to religion strikes a familiar
chord.  It seems we've talked about this topic several times over the
last 2-3 years on this and other nets.

I agree with David (up to a point); i.e., like religion, once you "see
the light" about quality, you don't understand why everybody doesn't
see it.  Like religion, you either accept the quality movement or
you don't--not exactly blind faith, but close.  Then, too, not
everyone wants or is able to make the leap of faith required to
accept the quality movement.  The more you think about it, the
more the quality "thing" seems like religion.  Amazing!

Tom Harris

Tom Harris, Ph. D., Prof. Bus. Mgt.
St. Philip's College
1801 M. L. King Blvd.
San Antonio, TX 78231
Voice 210-531-3494
"Continual improvement involving everyone"
...Imai

=====================================================================

Re: Lessons from Religion/Taurman

ILX

Thu, 10 Oct 1996 12:18:15 -0500 (CDT)

>                 David 

You ask for expansion on parrallel between religion and Deming. 
>
>
>==========================================================================

There are more analogies. I use some of them in the right crowd.

SPC is a core part of Demings ideas and drove him to develop a broad
understanding of management.

SPC is like Christ. Christ set the perfect example and asked us to follow.
But he also said forgive those who fall off the path and keep trying.

Managing with SPC is similar. SPC sets a goal of perfect no variation and
always in the center. It can not be done but managers who successfully use
SPC keep rewarding people for trying to stay on the path and reduce
variation around the target.

Another parallel. It takes crusade to change from the old ways to the new.
Only when there is a SPC or Deming zealot or crusader does culture change
happen. If there is doubt in the leader it spreads One must believe.

Another parallel is the Unknown. It is essential to proceed based on faith
that the change will be better. If there are doubt the change does not
happen but falters. It is not possible to prove beyond the shadow of a doubt
that the new Deming way will be better.

The worship of idols is hard to tell from the worship of God. It may look
the same it may sound the same but inside there is a difference. Deming
ideas sound similar to old ideas. Measure parts and record the data. In the
old system the data was used to tell good from bad. In the new system the
data is used to take action to fix and track progress. It looks the same but
there must be a difference in the heart of the manager for it to work  

Gene 

Eugene Taurman
interLinx
ilx@execpc.com
http://www.execpc.com/~ilx

=================================================================

Re: W. Edward's Deming Medal Nominations/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

11 Oct 96 10:53:00 EDT

The facts about the beginnings of the Deming Prize in Japan,  as related by Dr.
Deming are these:

He gave lectures in 1950,  setting forth for the Japanese his vision of what
good management is like.  There were stenographers present who took down what he
said and the materials were put together as a book.   After to book began to
sell quite well,  the Japanese came to Dr. Deming and proffered royalties.  He
refused and told them to take the money and use it to sponsor a prize for
quality.

I'm not sure what Matt means by "calibration of judges".  In both the Deming
Prize in Japan and in the USA,  judges are subjected to training in which they
practice rating various documents.  Scores are prepared and discussions are held
to reduce variation.   Over time,  as I understand it,  the variation among
judges is not very great.   When it is,  they look to find the special cause.

When the Baldrige prize was beginning,  I went over the procedures with my
teacher from Japan,  Professor Tsuda,  and then I passed his comments on to the
Baldrige office.  Curt Reimann,  who ran the office at that time,  wrote me a
nice letter discussing the ways in which he had dealt with the issues thus
raised.

Insofar as the Baldrige prize is concerned,  I certainly would not fault either
the procedures or the administration.   There are some differences between the
Deming Prize and the Baldrige prize,  having to do,  mostly,  with the fact that
the Deming prize is entirely a private sector matter whereas the Baldrige prize
was spawned in a political arena.   The developers of the procedures had to
watch out for certain constraints,  i.e.,  making the process Congress proof,
making it ecumencal so it did not favor one 'guru' over another,  making it
'objective' so that it would be hard to start a law suit,  etc., etc.

I was not present but I heard that Dr. Deming praised the folks who ran the
prize while at the same time criticizing its philosophical basis.   In Japan
there are winners,  but no losers.

Myron Tribus

====================================================================

Re: Postings and special cause(s)/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

11 Oct 96 10:53:04 EDT

This is a fun topic.  One thing I learned from Ed Jaynes,  who introduced me to
the maximum entropy estimate in probability, is this: Whenever you solve a
problem in probability,  people do not linger over your solution but rather they
criticize you for the problem you did not solve.

As to the Normal or Gaussian distribution:  I have been told that physicists
believe in it because they think the mathematicians have demonstrated it
describes so many things and mathematicians believe in it because they think the
experimentalists have 'proven' its worth.

Those of us who come at this issue from the perspective of the theory of
information take a different view:  To us,  probability is a number assigned to
allow us to tell someone else neither more nor less than we know about the
answer to a well defined question.  (Hence the insistence on operational
definitions.)

According to this perspective,  the function called 'entropy',  as defined by
Claude Shannon,  measures what we do NOT know when all we know is a probability
distribution.  If all we know is the mean and mean square of a distribution,
then maximum entropy will tell us that we should attempt to describe the system
by means of a Gaussian.   Note,  this is not the same as saying that the "real
distribution out there is a Gaussian".  Rather it is saying:  "This is the best
way I know how to describe the system,  given this information.  More
information may cause me to change my description,  provided it is inconsistent
with what you just told me."

Tonight,  after reading a few of the postings on this subject,  I took
Shewhart's book of the shelf and looked again at his derivation.   Maybe I am
biased (maybe???) but it seems to me that what he said in 1931 squares with what
Ed Jaynes taught me in 1958.   It strikes me as predictable that both Shewhart
and Shewhart would be products of the Bell Labs.  

Myron Tribus,  350 Britto Terrace,  Fremont, CA 94539
Ph:510 651 3641  Fax: 510 656 9875
I find that when I soak up information like a blotter
I often get it backwards.

=======================================================================

Re: Use of Surveys/Taurman

ILX

Fri, 11 Oct 1996 09:57:17 -0500 (CDT)

There are a variety of reasons why you and I have not seen improvement from
the surveys. Most of them are addressed in the book called the Trust Factor.
However I will try to explain my experience with surveys. 

Managers simply do not accept responsibility for worker attitude and do not
realize how long it takes to change an attitude once it is in the culture. I
have seen attitude abut rates persist for 10 years after rates were remove
form management list of measures. I have seen attitude caused by poor use of
performance review hang on for 10 years. Managers must drive the new culture
hard if they want to impact attitude. Simply stopping the old practice is
not enough. It must be driven out or the old attitudes caused by past
manager behavior will prevail. I have even seen attitudes developed in one
comapny brought to another. 

Managers most often believe attitude is important and feel that workers have
an attitude. How often have you heard an manager say, 'If only attitude were
better.' or 'If only they cared.' or 'why don't they care like I do?'. Is
important and speak of worker attitude as if he or she had nothing to with it.

The problem is they do not accept responsibility for attitude and realize
the manager actions cause attitude. So they take the survey but don't do
anything.

Managers fail to recognize the attitude and behavior are learn by workers
form the company consequence system and are passed on by peers. Employees
watch very carefully their peers and adopt the behavior that will minimize
their grief and maximize their reward. Each person has different definitions
of grief and reward based on their persona l values so same situation causes
different behavior.

The consequence system is broad. It include thank yous, put downs, staff
meeting consequences, promotions,  peer pressure, what managers choose to
fix what mangers do not fix etc.etc.

If people are subjected to scheduling systems or machine that do not wok
well they will develop attitude that mangers do not care about their time or
they are supposed to make do or it is more economical to limp along than fix
it right.

Managers do not seem to realize that their decisions cause attitude and
their main responsibility is to develop attitude. Until they understand how
and accept the responsibility attitude will not improve because of a survey.

I have written several articles trying to explain this phenomena. If you
want to read them I will send them. I would appreciate feed back on how to
improve  them.

Gene 

Eugene Taurman
interLinx
ilx@execpc.com
http://www.execpc.com/~ilx

======================================================================

Re: Postings and special cause(s)/Kromkowski

Kromkowski@aol.com

Fri, 11 Oct 1996 12:27:00 -0400

In a message dated 96-10-11 04:04:07 EDT, you someone wrote:

<< If October's data shows regression to the mean and falls back, then 
 the high September number is random, common cause variation.  If 
 however, October's data is also close to the 3 sigma + side, and so 
 is November's, then we could begin to consider >>

This may be a trivial point, but it has been suggested by students of the
philosophy of science that sometimes in a transition from one paradigm (<-
 this word should only be used if you have read Kuhn, and can at least
identify with some alacrity (accuracy is optional) where the book is in your
office, home  --  I think its on the bookshelf near my bed) to another,
people will use the language of the new model but in a way that is really the
old model.  In some respects, I have seen and perhaps even fallen myself into
the same trap, of using a control chart but then on future data     slipping
into a varient of the "action on the last data point" mode.

Putting aside any mathematical errors, if a point which falls out of control,
one may if economical (another key point about cc as a new "economic" tool)
decide to search for a special cause.  I did -- vacations in August.
 Confident in my explanation, I decided no further time should be wasted on
issue.  At that point, the control limits should be recalculated to exclude
"explained" anommoly.  

AFTER that, I personally don't think any more calculation of cc should be
done UNTIL the next special cause arises.  Hence, for future data points, it
is simply a matter of plotting point.  If within limits, move on to other
work. (Although, important, plotting and calculating charts produces no
product or service nor changes for better the attributes (<-- n.b. specific
decision to not use word quality, since I'm sure where my copy of Aristotle's
Politics is) of product or service.  I recall D writing about "profusion of
charts".

ONLY where future data actual signals something (rules for grouping, trends
should be used sparingly in my opinion, but then I'm still a hack) should
time be wasted on what could amount to a kind of 100% inspection of a stable
system.

The danger as I see it is that if one gets involved in constant analysis and
recalculation we're not doing prediction (i.e. into the future) anymore.

Naturally, if you don't like the sound of the stable system then DO something
about it . But as Yogi Bera may have said:  doing nothing is doing something.

JDKromkowski

=====================================================================

Re: Deployment Flowcharting/Clarke

"Don Clarke"

Fri, 11 Oct 1996 16:02:17 -0600

I also recommend the videotapes on Deployment Flow Charting (DFC) that
Myron Tribus recommends (they feature him, BTW).  I've used DFC a number of
times and find it much more enlightening than regular flowcharting.

DFC integrates the usual, impersonal flowcharting with the people
responsible for each task on the flowchart, which tends to perk those
people up.

==========================================================================

NASA, TQM, and SPK/Kerr

"Kerr, Donald"

Fri, 11 Oct 1996 17:05:00 -0500

Written in The NASA Systems Engineering Handbook is the following:

>Total Quality Management is the application of systems engineering to
the work >environment.  That is, part of the TQM paradigm is the
realization that an operating >organization is a particular kind of
system and should be engineered as one.  A >variety of specialized
tools have been developed for this application area: many of >them can
be recognized as established systems engineering tools, but with
different >names.  The injunction to focus on the satisfaction of
customer needs, for example, >is even expressed in similar terms.  The
use of statistical process control is akin to >the use of technical
performance and earned value measurements.  Quality >Function
Deployment is a technique of requirements analysis."

I'm a systems engineer. I've been working for NASA at the Johnson Space
Center for about 10 years now.  The above is very true.  Systems
engineering and earned value have made it into the dominant NASA
paradigm.  TQM has not.  [I must qualify that judgment.  I do NOT equate
TQM to the Deming Philosophy and Profound Knowledge.]  TQM as a
"program"  failed at the Johnson Space Center.  It was considered a
waste of time or not applicable.  I know, because I chaired one of their
Q+ Steering Committees.  Top Management support eroded over time and the
program merged with the employee suggestion process.  I once asked the
Director of Mission Operations (after I found out he attended the same
Deming 4-Day as I did), "Do you ever think the Deming philosophy will
ever make it at NASA?"  "Never!", he said.

 I worked Shuttle Missions as a Flight Controller.  The Flight Control
Room is full of people who spent two years planning for the flight, they
love what they do, they work together, they know the aim, they know how
they fit  in.  They understand clearly when to take corrective action
and when not to.  They know the whole is greater than the sum of their
parts.  It is the CAPCOMs job to talk to the crew.  It is everyone's job
to accomplish the aim and return the crew and the vehicle home.    They
understand intrinsic motivation (trust me, they are sure not there for
the paycheck!)  I know they understands systems, psychology, theory of
knowledge, and variation in extraordinary ways with extraordinary
results.  I even believe, in that room, on the day of flight, they
understand Profound Knowledge!

This still amazes me to no end. I'm perplexed!  If, according to NASA,
TQM is the same as systems engineering, why is the Deming Philosophy not
seen as related to "The Mission Control Team" philosophy and is seen as
an impossibility (Never!) to become the dominant paradigm at NASA?  
When the flight is over and the Flight Controllers return to their
offices, something is missing.  The "implicit" Profound Knowledge that
brought them such great joy...was left in the Flight Control Room. 
Deming, Bill Scherkenbach and others (including myself) have tried to
make Profound Knowledge "explicit"  to NASA management on several
occassions.  Why can they not make the connections stated above as I
have?

My theory is, beyond blaming paradigms, is the way the Deming  message
has been taught. The predominant emphasis on  statistics and performance
appraisals causes the NASA community to make a leap of abstraction
"Doesn't apply here...and the government requires
appraisals!...Communism!..."  They seem to shut down at this point. 
After that the mere mention of the name Deming conjures up these images.

I said all that to say this:  The "rules" in the Deming community must
change.  We need to better define Profound Knowledge so that is does not
get replaced my misunderstanding such as TQM, Systems Engineering,
Communism, etc.  We need to make all the implicit assumptions and
beliefs behind the Deming Philosophy more explicit, without relying so
much on statistics,Red Beads, 14 points, etc.  Profound Knowledge is so
profoundly simple, as the teaching's of Jesus are.  Religion, or the
teachings about Jesus, has made it more complicated to communicate the
Good News.  Have we done the same with Deming?

Have a Great Adventure!
Don Kerr
Donald.Kerr@USAHQ.UnitedSpaceAlliance.com

============================================================================

Re: Use of Surveys/Tolman

"Anton O. Tolman, Ph.D."

Fri, 11 Oct 1996 17:09:20 MST

On October 10, Steven Prevette wrote:

>      The question to DEN members is:  does anyone have experience with the 
>      use of personnel survey results as part of an application of Deming 
>      management theory, and more importantly, an application that lead to a 
>      significant improvement in business success?


I hope I don't disappoint you Steven when I say that I am not going 
to directly answer your question.  I thought your comments were on 
target as to why personnel surveys often produce no results -- they 
are based on the assumption that the worker is at fault.  What if the 
HR department knew/understood that the employees were the customers?  
After all, without employees, there are no personnel to worry about.  

On the other hand, I believe that Deming was a strong proponent of 
__customer__ surveys -- in attempting to understand your customer, 
you evaluate the customer's attitudes and responses toward the 
organization; that allows you to prioritize improvement projects and 
keep the customer focus in mind.  The emphasis on suppliers likewise 
enables the organization to improve its processes overall.  

I'm eager to see the replies to your query.  
________________________________
Anton O. Tolman, Ph.D.
Director, Psychological Services
Director, Quality Management
Wyoming State Hospital
P.O. Box 177
Evanston, WY  82931-0177
Anton@wsh.state.wy.us
(307) 789-3464
---------------    

===========================================================================

Re: Customer expectations/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

12 Oct 96 00:16:21 EDT

Kris wrote:

>Quality is an ambiguous term at best.  It means so many different things
>and can be defined so many different ways as to be virtually meaningless.

IMHO the phrase "virtually meaningless" isn't quite right.  As Persig said, in
his wonderful book,  "Zen and the Art of Motorcycle Maintenance",  quality is a
primitive.   So is "love".  Love is very difficult to define.  That is because
it is a basic, primitive idea.  You cannot define primitives in terms of more
primitive concepts.  

Love, too,  is hard to define.   But that does not make it a meaningless
concept.

To make quality an operational concept requires that we define the situation in
which we intend to apply the concept.   The concept 'right answer' is useless
unless the question is defined.   Quality is not a useful concept until we find
ourselves discussing a product or a service in a particular context.   Then it
is possible to develop the attributes which bear upon quality and,  then,  in
that context,  give an operational definition FOR THE MOMENT.  Ideas about what
constitutes quality change with experience.   But then my idea of the woman I
want to love at age 75 is quite different than my ideal at age 24.

Myron Tribus,  350 Britto Terrace,  Fremont, CA 94539
Ph:510 651 3641  Fax: 510 656 9875
I find that when I soak up information like a blotter
I often get it backwards.

==========================================================================

Re: Systemic -vs- systematic/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

Sat, 12 Oct 96 01:16:24 EDT

Deployment flow charting is a special form of flow charting.  At the top of the
page you put the "cast of characters"...that is,  the names of all the
significant players in the work.   Their names are in boxes across the top of
the page.  Then,  whenever a task is to be accomplished,   you draw the flow
chart symbol (report, task, decision, meeting,  document,  assist,  etc.) under
the person most responsible.  Then when the chart is finished you will not only
see what has to be done,  but you will also see the horizontal lines showing
when the work passed from one person to another.   These horizontal lines define
customer-supplier relatinships.

There is a very nifty set of software for this purpose.  A bit expensive,  but
designed especially for deployment flow charting.   It is called "Team Flow" and
sold by CFM, Inc. Sixty the Great Road, P.O. Box 353,  Bedford,  MA 01730-0353
(617) 275 5258   FAX: (617) 275 7008   e-mail:  CFMpro@applelink.apple.com

Ron Cordes,  who wrote the program,  and founded CFM, told me he produced it
because he found the materials in my video tapes helpful.  These tapes,
"Deployment Flow Charting" vols I and II are available with workbooks from
Quality and Productivity,  Inc., 1081 Westwood Blvd, Suite 214,  Los Angeles  CA
90024.  (310) 824 9623   FAX: (310) 824 7170.

Examples of flow charts are in several of my publications.  They are collected
under the title "Quality First",  available from the National Society of
Professional Engineers,  1420 King Street, Alexandria, VA  22314   (703) 684
2800
FAX:  (703) 836 4875   Ask for publication #1459.   The papers therein are NOT
copyrighted.  You can copy them as you wish.

Most of the papers in that collection have been placed in the archives of both
the DEN List and the CQEN.  I don't know precisely how to get at them but a call
to the den master should get you the information.   The collected papers sell
for under $20 so you can trade that against the time you spend downloading,
printing, etc.


Policy Deployment is an entirely different topic.  The word deployment shows up
on each topic because the concern is with deploying something (a policy or the
efforts of people) in an enterprise.

The simplest definition I can give is this:  Instead of top down, firm command
and control,  which characterizes the dream of old fashioned management (and the
nightmare for the people in the system) quality management depends upon decision
making at the lowest possible level.  But people cannot be allowed to just do
what they wish.   There has to be some kind of coordination.  This is to be
achieved by the issuance of policy guidance.  We are not speaking of policies
that are written in a book somewhere and then forgotten.  Rather we are talking
about a method of deploying policies throughout a company and then following up
to find out if they help anybody.  Thus there is a special system created with
these features:

General direction from the top  (This year we are going to emphasize model x's
because that's what the customer studies predict)
REspons from the lower levels:  (These are the things we intend to do to help
implement the policy)
Multiple meetings to locate bottlenecks and places where the system has to
improve.  Discussions,  up and down the line,  regarding how to take advantage
of initiatives.

As a result of this process,  people know what to do because they proposed to do
it and it was seen,  through consensus building,  to be a good thing to do, to
be within the capacity of the staff, etc.

There's a lot more to it than that.  i once prepared a manual for jaguar on how
to carry out policy deployment,  but Ford bought jaguar and I became an
unnecessary wheel on that vehicle.  I believe someone in the UK has undertaken
to publish it,  but i have lost track.

You see,  i no longer care about the publish or perish routine.  I expect to
perish anyway.

Myron Tribus

=====================================================================

Deming Dialog and Trademarks/Clauson

clauson@deming.ces.clemson.edu (List Moderator Account)

Sat, 12 Oct 1996 02:51:26 -0400 (EDT)

At today's Institute Design Council, the following were approved:

1.  Based on the DEN survey, the in-process Institute newsletter will be named 
	The Deming Dialog.

2.  A trademark has been issued for:
	The W. Edwards Deming Institute (tm)

3.  A trademark has been issued for:
	The Deming System of Profound Knowledge (tm)

Jim Clauson
DEN Moderator

======================================================================

-- 

Deming & JUSE in 1950's/Latzko

"William J. Latzko"

Fri, 11 Oct 1996 23:25:24 +0000

[Bill is responding to Matt's question about the early work of Dr. Deming
in Japan and the monies used to start the Deming Prize.]

The Union of Japanese Scientists and Engineers (JUSE) published Dr. Deming's
lectures in a book called ELEMENTARY PRINCIPLES OF STATISTICAL CONTROL OF
QUALITY. This book sold very well in Japan. When JUSE wanted to give the
royalties to Dr. Deming, he declined them and asked that they be used to
promote quality in Japanese products. JUSE used the funds to start the
Deming Prize.

I have the good fortune to have a copy of the second edition of this book.
The editor's preface has much background about Dr. Deming's trip to Japan in
1950. Since the book is hard to get, I thought that the DEN members might
gain some insight from a firsthand account of how Dr. Deming's lectures in
Japan came about. I use asterisks (*) in lieu of quotation marks to
accommodate those systems that cannot use the latter.

Bill Latzko

---------------------EDITOR'S PREFACE---------------------------------------

	This is a shorthand transcript of Dr. W. Edwards Deming's lectures on
*Statistical Control of Quality.*
	As everybody knows, Dr. Deming is one of the world-famous authorities on
the theory of sampling.  In the field of statistical quality control, i. e.
application to industrial production of mathematical statistics, too, he is
versed in both theory and practice.  This is clear from the fact that he has
published many papers and books, especially notable his *Some Theory of
Sampling*.  It was he who edited W. A. Shewhart's *Statistical Method from
Viewpoint of Quality Control* (1939).  As a member of the five-man Emergency
Technical Committee, he played  a major role, in collaboration with Messrs.
H. F. Dodge, A.G. Ashcroft, Leslie E. Simon, R.E. Wareham and John Gaillard,
in the compilation of the American War Standards (ASA ZI.1-3). He has since
been striving in earnest for wider application of this method; and Mr. W. B.
Rice and other experts in quality control dedicate their books to Dr. Deming
(for instance, Rice's *Control Charts in Factory Management*, 1947).
	Therefore, Dr. Deming had been the idol of Japanese students in quality
control.  Fortunately, when he revisited Japan in the summer of 1950, he
kindly accepted our request for a lecture course and gave a series of
lectures on statistical control of quality for eight days (10-18 July) at
the Nippon Ishikaikan Hall, Kanda, Ochanomizu, Tokyo, to 230 first-class
engineers of leading manufacturing corporations.  His lectures were of the
same standard as those he gave in the United States where statistical
quality control methods have made remarkable progress since wartime.
Throughout the 8-day course there always prevailed an atmosphere of unity
between the teacher and students.  With simple explanations and adequate
demonstrations, Dr. Deming's lectures were so effective and persuasive that
they left an unforgettable impression upon our minds.
	With a view to attaining satisfactory results in the projected lecture
course, the Union of Japanese Scientists and Engineers had made careful
preparations:
	1)	Mr. Hisamichi Kano was selected is interpreter.  As Dr. Deming himself
recognized with thanks, Mr.  Kano is an interpreter of rare talent and
knowledge in English.
	2)	Many Japanese scientists and GHQ officials were so kind as to make
comments on Dr. Deming's lectures or deliver supplementary lectures.  They
were, among others: Dr. Motosaburo Masuyama, Dr. Eizaburo Nishibori, Dr.
Shigeru Mizuno, Mr. Heihachi Sakamoto, Mr. H. M. Sarasohn, Mr. W. MacPhee,
Mr. Kaoru Ishikawa, Mr. Masao Kogure, Mr. Shin Miura, Mr. Eizo Watanabe, and
Mr. Masao Goto.
	3)	 Mr. Susumu Sonobe, of the Nippon Denki K. K., volunteered to take care
of the public address system in the Lecture Hall.
	4)	In order to make a perfect record of Dr. Deming's lectures, a recording
subcommittee was formed by Dr. Mizuno (as chairman), Mr. Ikuro Kusaba, Mr.
Tadasu Fujita and Mr. Masumasa lmaizumi.  Both Japanese and English
stenographers were employed, We thought it essential to make a perfect
record of Dr. Deming's authoritative lectures because there had been in
Japan some sort of confusion and disorder in the application to industry of
statistical quality control methods, which we hoped might be brought to a
settlement by Dr. Deming's enlightenment; and inasmuch as such a record
would be of great use not only to the students who attended the course but
also to many others in manufacturing plants for repeated study of Dr
Deming's lectures and the development of new ideas and devices.
Fortunately, Dr. Deming allowed us to publish it in book form.
	After the lecture course was over, Mr. Kano carefully edited the English
shorthand record.  In the case of the Japanese edition, the members of the
recording subcommittee lavished efforts, for theoretical explanations,
mathematical formulas, unification of technical terms, etc.  Efforts were
made also to effect a unity between the Japanese and English editions.
	Not until about three months after Dr. Deming's departure for the United
States was our editing work finished.  We deeply regret that the English
shorthand record was not supervised by Dr. Deming himself, but we think
that, thanks to, the unstinted efforts on the part of our experts, the
record of Dr. Deming's authoritative lectures has been brought to the point
of perfection.  We are looking forward that Dr. Deming's lectures (the
flowing, elegant English edition and the correctly-translated,
carefully-worded Japanese edition) will find a large circle of readers among
our managers and engineers, and exercise far-reaching influence on the
application in Japan of statistical quality control methods, as Lafcadio
Hearn's (1850-1904) *Lectures on Literature* were read widely by Japanese
writers and authors to give inspiration to the literary world here.
We should like to explain in brief the circumstances in which Dr. Deming
kindly consented to give his lectures above explained.  In December, 1949,
Mr. Sigeiti Moriguti, of Tokyo University, received a letter from Dr.
Deming, informing that he would come to Japan in the spring of 1950.  On 8
March 1950 we wrote a letter to Dr. Deming, asking that a message be sent to
the first issue of our organ *Hinshitsu Kanri* (Statistical Quality Control)
for which preparations had almost been completed, and inquiring after the
exact date of his arrival here and itinerary.  Incidentally, we had been
continuing our research works on statistical quality control since 1948 and
were holding the SQC Seminar, 30 days a year, attended by 39 engineers of 17
corporations.  And our research had developed to the point where we could
publish a monthly organ, *Statistical Quality Control.*
	On 22 March we received a kind letter from Dr. Deming, saying that he would
visit this country in the summer of 1950, and asking what kind of article we
would like to have.  He also recommended the following American scholars to
whom our monthly organ should be sent with compliments:
	Charles A. Bicking             Dr. Martin A. Brumbaugh
	Harold A. Dodge	               Dr. E. J. Gumbel
	Howard L. Jones	               Paul Peach
	David H. Schwartz              Dr. Water A. Shewhart
	Ralph E. Wareham               Charles D. Wiman
	Prof. Mason E. Westcott        Major-Gen. L. E. Simon
	Prof. Lloyd A. Knowler

	Deeply moved by the kind letter, on 1 April we wrote again to Dr. Deming,
informing him of the title and content of the article which we would like to
have.  In the same letter, we also asked if he could give a series of
lectures on statistical quality control for several days while he was in
Tokyo, for this plan had been readily approved by our president, Ichiro
Ishikawa, and our lecturers who had been cooperating with us in the SQC Seminar.
	We soon received a letter dated 11 April from his secretary saying that Dr.
Deming was traveling at that time.  Then, a reply of 18 April addressed to
Managing Director K. Koyanagi was forwarded to our office from Miss Margaret
Stone, PSD, ESS, SCAP.  Several days after, an air mail letter dated 22
April almost to the same effect reached us.  Enclosed therewith we found *an
outline of course in sampling*. In these two letters, Dr. Deming kindly
accepted our proposal for a lecture course, saying as follows:
	*I am deeply honored by your invitation to give some lectures this summer
and to assist with your proposed course in the statistical theory of control
of quality and its applications.  First of all let me say that your plan for
an intensive course lasting 5 to 7 days, 7 hours per day, is
excellent........ It will be possible, I am sure, for me to assist you with
this course by giving some lectures on statistical theory and demonstrations
of applications, from 2 to 4 hours daily........  In this country we found
that 8 days was a minimum for people who had not studied elementary
statistical theory. I should think that in Tokyo the 7--day period  would
probably be much better than the 5-day period.  If possible, you might even
extend the time to 8 days. - - - - -  *
	Thereafter, Dr. Deming accorded us all sorts of facility to make effective
the projected course.  For instance, he recommended as a textbook the
*Control Chart Method of Controlling *Quality During Production (ASA ZI.3)
and sent us six copies of it; he brought with him equipment necessary for
demonstrations of applications; though he was extremely busy in Tokyo, he
wrote  us a draft for an advanced course; and, above all, he kindly
recommended us to Mr. Kenneth Morrow, Chief of PSD, GHQ, SCAP, who helped us
in every way to conduct the lecture course with success.  Thus, the success
of the course was for the most part to the credit of Dr. Deming.
	Under these circumstances and thanks to the good offices and cooperation
not only of Dr. Deming but also many friends of ours, this transcript of
lectures on statistical control of quality has come to be published after
having obtained the lecturer's approval.  We take this occasion to express
our heartfelt thanks to all of them, particularly Mr. Kenneth Morrow, the
Japanese scholars who willingly assisted us, and Mr. Shoichiro Niki and
other members of our secretariat.
	10 November 1950
	Ken-ichi Koyanagi
	Managing Director, Union of
	Japanese Scientists & Engineers
---------------------------END OF EDITOR'S PREFACE-----------------------------
---------------------------------------------------------------------------
William J. Latzko
215 - 79th Street
N. Bergen, NJ 07047
Voice:     201-868-5338
Facsimile: 201-868-5338
E-mail:    latzko@worldnet.att.net
------------------------------------------------------------------------------

=======================================================================

Deming Questions # 30/Gogue

gogue@cri.ensmp.fr (GOGUE J.M. Societe MAST 39 50 99 67)

Sat, 12 Oct 96 09:54:22 +0200

--------------------------  DEMING QUESTIONS  -------------------------- 

A weekly paper including a set of questions copied in the Deming book 
OUT OF THE CRISIS   Chapter 5.  Questions to Help Managers.
By Jean-Marie Gogue

-----

30-1.   (No 59 in the Book) - Management by figures

a)     Do you run your company on visible figures alone?

b)     If yes, why?

c)     What steps are your management taking to learn the importance 
         of figures unknown and unknowable?

COMMENTS
The great majority of people believe that a company MUST be run on 
visible figures alone. Why? They want a scientific management -- right  
way -- but they believe that "Scientific" means "Based on Figures Only" 
-- wrong. A good instance of a scientific approach that is not based on 
figures alone is modern medecine. Perhaps many diseases could be explained 
better with unknown figures. Visible figures (e.g. pulse) can help however.
  
-----

30-2.   (No 60 in the Book) - Involvement in Standard Organizations

Does your company participate in committees of standardizing bodies?

COMMENTS
Standard committees are good sources of knowledge. The late lamented 
Dr. Kaoru Ishikawa was involved in JIS (Japanese Industrial Standards) 
and ISO committees. He wrote: "in implementing Quality Control in 
various industries, it became clear that the existing Japanese Industrial 
Standards were not adequate or accurate".

-----

Jean-Marie Gogue
President
The French Deming Association
Versailles   France
gogue@ensmp.fr

===========================================================================

Re: Deployment Flowcharting/Baxter

baxter@rsc.co.uk (nick baxter)

Sat, 12 Oct 1996 09:20:08 +0100

>Myron
You mention software but dont give the reference. I would be grateful for
further info on this. The BDA also has a nice little booklet on flowcharting.
Thanks also for the reference through the Den 'Global Quality'. I have got
it through the British Lending Library service who found a copy in
Portmouth. Quite a work!

===========================================================================

Re: Use of Surveys/ Kerridge

David Kerridge

Sat, 12 Oct 1996 11:44:20 +0100

Steve Prevette says:

>     I have been having an off-line discussion with Geoff Penrose about the
>     use of personnel surveys -> measuring worker attitude -> determining
>     common causes & special causes (as appropriate to the statistical
>     analysis of responses) in order to improve worker attitude, and
>     therefore business success.  I will admit that I have been trained in
>     survey construction and statistical analysis of survey results.
>     However, I have not (yet?) been involved in an application of survey
>     results (either of workers or customers) that seemed to actually cause
>     a significant improvement in business success.
>

A very good point. The same observation can be made about "Customer
Satisfaction Surveys". Most generate huge reports which are filed and
forgotten. This does not mean that studying customers is a waste of time:
it is an essential part of the "Production viewed as a system" flow
chart. But most surveys do not provide the kind of information which
enable you to act more effectively.

The only personnel survey I have seen was just like this: it was concerned
with *asking* people if they were satisfied with (a list of things). All
that could be concluded was that some were more satisfied than others, or
at any rate more likely to complain. This is just like customers: and, of
course, personnel are customers of management services.

If you want to predict action, you should study action, not words. A
formal survey may or may not be a good way to do this. Since in most
cases, especially within your own company, you want to know what to
*try*, in the PDSA cycle, rather than what large-scale decision to take,
you do not need the right answer, just an interesting theory to try.
Get a group together, and ideas about what to try (both good and bad)
will usually flow faster than you can keep up with.

Most surveys simply provide the wrong kind of information, and it is
just as well that they are ignored. I believe my paper "What do
Customers Want?" is available from the DEN Web site.

David Kerridge

British Deming Association Scotland
dfk@rsc.co.uk

===================================================================

Resources for Feuerstein materials/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

12 Oct 96 13:57:15 EDT

Resources for learning about Reuven Feuerstein's work.

In the USA the principal source of books, tapes and consultancy about
Instrumental Enrichment (IE),  Mediated learning experience (MLE) and the
Learning Propensity Assessment Device (LPAD) is an outfit called IRI/Skylight
(IRI = International Renewal Institute).   You can reach them at:

IRI/Skylight
Training and Publishing Company, Inc.
2626 S. Clearbrook Drive
Arlington Heights, IL 60005

Tel:  (800) 348 4474   or (847) 290 6600
FAX:  (847) 290 6609

They have an extensive catalog of materials on various approaches to improved
learning.  Among them are:

"On Feuerstein's Instrumental Enrichment"  a collection edited by Meir Ben-Hur.
($22.95).   In my opinion,  this is the best introduction to the method
available.  I tell beginners to start here.

"Changing Children's Minds" by Howard Sharron ($32.00)   Also a good
introduction and certainly the second book I would recommend.

"Mediated Leraning In and Out of the Classroom" a manual prepared by people from
Soth Africa to help teams who are introducing IE, MLE and LPAD.  ($29.95)    I
have not studied this manual.

"The Mind of a Child" a videotape around the story of Lorna Williams,  an
American Indian who found in Feuerstein's methods a way to rescue Indian
children who were failing in the White Man's system of education.   This is a
very moving story and has important lessons for our dealings with minorities in
our society.  

In earlier catalogs from IRI/Skylight,  I believe I saw some audio tapes in
which Dr. Feuerstein discusses some of his experiences.  These were also very
moving for me,  especially the story of how a young inner city boy tells why,
after using cocaine from the age of 12 to 14 he intends to give it up, "Because
I have discovered I have a brain."

I have some other books which are not available from IRI and belong in the
library of anyone who is going to go beyond the basic introduction.  These books
are:

"Don't Accept Me as I Am: Helping 'Retarded' People to Excel" by Reuven
Feuerstein, Yaacov Rand, John E. Rynders,  Plenum Press  (1988)  This book
discusses the philosophy,  the science,  the methods and the experiences using
MLE, IE and LPAD with children who are considered 'learning disabled' and
getting them to excel.  Most of the methods described in the book are directly
applicable to teaching those who are 'temporarily enabled'.  (We all become
disabled,  eventually.)

"Instrumental Enrichment:  An intervention Program for Cognitive Modifiability"
by Reuven Feuerstein and others.  This book is available from Scott,  Foresman
and Company,  1900 East Lake Avenue, Glenview, IL 60025.  It was out of print
for some time.  AFTER taking a course of instruction from Feuerstein,  I could
read this book quite easily.   Maybe it is my own limitation,  but before taking
his course I found the book difficult to read.

"Mediated Learning Experience (MLE),  Theoretical Psychosocial and Learning
Implications" by Reuven Feuerstein,  Pnina S. Klein and Abraham J. Tannenbaum.
Freund Publishing House, Ltd., Suite 500, Chesham House, 150 Regent Street,
London, W1R 5FA.    This book is not readily available from sources in the USA.
I understand that a revision is contemplated.   After you have read some of the
other books,  this will be found to be a very handy reference.

>From time to time the IRI/Skylight sponsors conferences,  seminars and training
sessions on MLE, IE and LPAD.  I can heartily endorse the experience.

Myron Tribus,  350 Britto Terrace,  Fremont, CA 94539
Ph:510 651 3641  Fax: 510 656 9875

==========================================================================

Re: Deployment Flowcharting/Wendt

John Wendt

Sat, 12 Oct 1996 16:50:07 -0500

I would like to "chime in" on deployment flow charts.  I believe they 
are very powerful when used correctly.  It, like many other improvment 
tools, is a "sleeper".  It will not fix anything on its own, and does 
not appear, initially at least, dynamic or dramatic.

Having gotten my initial start from Myron, I have also found that, since 
the deployment style causes the question of "Who is responsible for this 
activity, meeting, decision, etc." to come up frequently, it often 
forces clarity or flushes out issues that have not been attended to, and 
the creation is an education in itself.

Myron's comment, which I use whenever I teach this, is that, "You don't 
learn to chart so much as you chart to learn (the way things really 
are).  I have heard countless times the belief that "we shouldn't be 
constrained by the past--let's start with a clean sheet of paper!"  
While there are times for this, when trying to solve system performance 
problems, my belief/statement is that if you don't REALLY understand the 
current reality, you will only rearrange it or put a different coat of 
paint on it, but make little fundamental improvement.

The last time I used this approach at Procter & Gamble, it took seven 
(7) meetings and flow chart reviews before the group--only myself and 
two others--could agree that we had truly captured current state.  In 
the press of business, most organizations don't have this patience or 
discipline.

I don't mind if an organization starts with a simple flow chart, but 
caution them that we should not end there if they really want to make a 
difference.

======================================================================

Re: W. Edward's Deming Medal Nominations/Tribus

mbc@cosmail3.ctd.ornl.gov (Jack Campbell)

Mon, 14 Oct 1996 08:32:55 -0500

On 10/11/96 you wrote the following:

(SNIP)

.............
>I was not present but I heard that Dr. Deming praised the folks who ran the
>(Baldridge) prize while at the same time criticizing its philosophical
basis.   
...

>Myron Tribus

--------------------------------

Question:  In the debate regarding whether Deming was for or against awards,
the note about Deming "criticizing the philosophical basis" of the Baldridge
award seems significant.  Any idea about what he took objection to ? 

Regarding the Deming Medal, if everyone can win it, what are the 
qualifications.  Are they commonly known among people in the quality field ?

Jack Campbell
Cycle Time Reduction Facilitator

[Moderator's note: If you are asking about the JUSE-administered Deming Prize,
JUSE recently allowed us to send out the criteria via the DEN - which I
posted to the DEN and will (eventually) get on the Clemson web site.

OTOH - If you are asking about the AQSC-administered Deming Medal, Bill Latzko
recently posted a summary of the qualifications.
 
======================================================================





------- End of Forwarded Message

Re: NASA, TQM, and SPK/Newman

Downstrm@aol.com

Mon, 14 Oct 1996 12:39:06 -0400

I believe that we have made the Deming Philosophy far more complicated than
it needs to be.  A philosophy is not intended to be a detailed strict set of
rules and regulations that are to be followed and if you do then you win the
game.  It is intended to be a visionary guide to help people/companies focus
their thinking and activities to improvement.  It is a sounding wall or
foundation used to support the analytical processes we use to make decisions.
 The philosophy itself never provides the answer, rather it provides the
framework/processes to be able to develop the answer.  I think in attempting
to wrestle the answer from the philosophy people continue to try and mold the
philosophy into the answer.  This process begins to pile layer upon layer of
detail on the philosophy shrouding it's orginal simple form making it harder
and harder to communicate to new people.

Another comment about the Mission Control Team.  I have seen this same thing
at the last company I worked for.  Our business was seasonal.  We made
capital equipment so there was always a big demand for shipment prior to
christmas so it could be installed during holiday shutdown.  Our
manufacturing was divided into 5 departments which as a system was able to
ship on average 2 million $/month(verified in control charts).  Almost every
December we were able to ship significantly more work(charts showed special
causes) even though most years had fewer shipping days in that month. How
could they do it.  if you watch how the 5 departments worked together during
december it was amazing to see them work as an almost seamless process
stepping in to help when ever and where ever needed.  Offering each other
suggestions to improve their areas that normally would have been looked at as
if the person was trying to take control of another  area.  For one short
month they actually were a system and were living the philosophy, but they
would never acknowledge it.  Many times I tried to get the director of mfg.
to use the ususally slow january to try an analize how the system functioned
during december and incorporate some of those processes into the normal day
to day operations, but he did not see it as a deming issue.



Mike Newman
Downstrm@aol.com

========================================================================

Leavenworth and CLT/Look

"Look, Alson"

Mon, 14 Oct 1996 15:27:00 -0400

I had a chance to talk to Leavenworth back in 1991 about his comment on
the CLT.  He indicated that the real basis for the control chart was
that it works.  He said that the CLT makes it reasonable for us to do
things like OC curves, associate probabilities and so forth.
Later after the 6th edition but before the 7th edition I again had a
chance to talk Leavenworth about his comment on the CLT.  I pointed out
to him that what was written in the 6th edition may cause confusion
about the nature of the control chart.  Leavenworth indicated again that
the basis for the control chart was not the CLT and that he would make a
note on the subject.  Hence I believe that is what he attempted to do in
the 7th edition.

Alson C H Look
EJLA29B@prodigy.com

Standard disclaimers

===========================================================================

Re: Customer expectations/Tolman

"Anton O. Tolman, Ph.D."

Tue, 15 Oct 1996 14:37:39 MST

On Thursday, October 10th, Kris Carlson wrote:

> Dr. Tolman replied to Julie Beedon's post which mentioned her comment about
> Quality being what all the stakeholders say it is.  Dr. Tolman argued
> against that by using the example of a hospital with no customers but which
> made all the other stakeholders happy.  This is an example of reducing an
> argument to an absurd level, demolishing the absurdity, and then stating
> you have demolished the original point.  (This technique is frequently used
> in political campaigns, by the way.)  Julie's point is still valid, since
> no organization without customers can satisfy the remaining stakeholders.  

I may differ from Kris in that I think that it is important to work 
through the meaning of the term "quality".  As I am engaged in trying 
to teach others in my organization how to shift their thinking along 
these lines, the definition of what is quality work and productivity 
becomes important to define.  Most of the people I am trying to teach 
do not understand or, better said, do not share my understanding of 
what quality refers to in this setting.  In fact, most of them do not 
think of their customer when they go about their work.  To be honest, 
most of the employees in this organization think of the "boss" as 
their customer.  This causes many problems as we try to work our way 
through to actually improving the services we provide to patients, 
who are our true customer.  

My example of the hospital with no customers was not intended as a 
pure reductio ad absurdem arguement.  I think that in my state, we 
are currently in a system where the stakeholders (e.g. courts, mental 
health centers, and many others) would be more than content for us to 
remain a center for keeping disturbing people locked up for long 
periods of time.  Many of our legislators, who are certainly 
stakeholders in a state system, do not understand anything about 
mental illness, about the pain it causes, and are more concerned 
about grazing rights, mineral revenues, etc.  In this case, it 
becomes very important to this facility to determine who our customer 
is -- is it the patient to whom we should be providing high quality 
services that benefit not only that person, but ultimately the state? 
Or is the quality of our services defined by paying attention and 
redefining our stakeholders as our customers?  Certainly, the courts, 
the legislature, the CMHC's etc. are part of these system and we need 
to listen to them and establish working and effective relationships 
with them.  But the definition of whether we are producing quality 
bring us straight back, in my mind, to the patients.  
 
> Quality is an ambiguous term at best.  It means so many different things
> and can be defined so many different ways as to be virtually meaningless. 
> Deming himself defined quality both in terms of the customer and in terms
> of satisfied employees.  Both are valid, yet both do not have to be true at
> the same time.  A Swiss watchmaker may have had great pride in his work,
> but the customer demanded greater accuracy at lower cost.  Digital watches
> made proud Swiss watchmakers obsolete.  

While I agree with Kris that "quality" may be an ambiguous term, I 
still think that for the organization's own purposes and for 
achieving constancy of purpose, that it needs to be defined within 
that organization.  Without this, there is no constancy of purpose.  
I like the reference to the Swiss watch industry and remember a Joel 
Barker film where he discusses this in terms of the importance of 
vision and recognizing opportunities to see the future.  The point 
Barker made with the Swiss was, if I remember correctly, that one of 
their own came up with the idea for the digital watches, but was 
rejected by the industry due to a very narrow focus or definition of 
a watch.  In other words, digitals were not watches and the Swiss 
were in the business of making "high quality" watches.  However, when 
the digitals came out, the customers flocked to them.  The lesson?  
The Swiss were more interested in  own vision of watch-making 
that they forgot to see or understand their 's vision of a 
watch.  This reinforces the point I made above.  In order to survive, 
companies must be able to integrate and walk the tightrope between 
intense focus on quality outcomes to customers and innovation to meet 
the future needs of customers.  I believe Deming often stated that we 
should aim to __exceed__ the customer's expectations.  

________________________________
Anton O. Tolman, Ph.D.
Director, Psychological Services
Director, Quality Management
Wyoming State Hospital
P.O. Box 177
Evanston, WY  82931-0177
Anton@wsh.state.wy.us
(307) 789-3464
---------------    

[Moderator's note:  Before we repeat what I have seen on other lists - could 
I ask that all 600 +/- subscribers *NOT* reply with their own definition of
quality.  Anton's point about an organizational context for the definition
is quite relevant to this.  Thanks.                     Jim Clauson]

Where's Clauson?/Clauson

clauson@deming.ces.clemson.edu (List Moderator Account)

Tue, 15 Oct 1996 22:00:51 -0400 (EDT)

DENizens,

In case you are wondering about the quiet period on the DEN...

This past weekend was the 6th Deming Institute conference in DC.  I had a 
"laptop" with me (386, 2400 baud modem) which took an inordinate amount of time
to process messages Thursday and Friday nites after the conference - so I 
"took off" Sunday and Monday nites -- but will get caught up tonite.

For those of you that could not attend - I'll be passing along highlights as
well as new trainer's exchange exercises.

Jim Clauson
DEN Moderator

=============================================================

Re: Leadership, knowing & Plato's Cave/Stauffer

Ripstaur@aol.com

Wed, 16 Oct 1996 05:54:03 -0400

In a message dated 96-10-03 23:13:33 EDT, Mike Newman wrote:

<< Now here is the question.  How does one extricate themselves from this
 situation.  if you continue to focus on top management you risk further
 isolation and increase their view that you are out of touch and not
 supporting the organization.  Or...  Soften your approach at the top, but
 lose credibility with the rest the organization and maybe yourself because
 you aren't fighting the battle they/you know you should be.  In either of
 these cases you may be of no future value to the organization.  On the first
 hand you might find yourself no longer with the organization and on the
 second hand could result in you being in a lame duck position seen as one
not
 willing to walk the talk by the rest of the organization.
  >>

I don't know what civilian CEOs are like, but from what I read in this forum
and other publications, they are much like the top military leadership I deal
with. It has been my experience that people at the top are much like everyone
else, but their isolation leaves them, for the most part, ignorant of the
information they most need to do their jobs. People who work for them have
(for a variety of reasons) been telling them "what they want to hear" for so
long they no longer have any idea what is really going on in the
organization. 

Good leaders, I mean ones that are inherently trying to do the right thing,
will appreciate someone who will tell them the truth. They are grateful to be
told they have no clothes (especially if you can tell them BEFORE they get to
the parade!)

There has been a lot of discussion in government circles recently about
"Stealth TQ." The idea is, just use the tools and principles, but never
mention Deming or Total Quality. That way, you don't run the risk of
"offending anybody" or "turning them off." I have seen this idea, couched in
other language, in recent postings on this forum. 

Well, I don't know about anyone else out there, but I refuse to turn this
into some sort of Amway-type sales scheme. I don't think you can build trust
with anyone if you don't show your committment up front. Most people, when
truly confronted with the ideas in the SoPK, find them extremely hard to
argue with. Most of their arguments are with perceptions they have
erroneously picked up somewhere along the line, and these arguments crumble
in the face of real knowledge.

Of course, some myths will live on for a while. Pay for performance is pretty
well ingrained, as are preformance appraisals. People will continue to use
specification limits as some kind of gauge for achieving quality, and will
continue to believe that there is some objective true value for everything.
These paradigms are tough to break, but usually, if you can kill one, the
rest fall like dominos. This, I think, is the strength of the SoPK. Once you
begin to understand it, it becomes almost impossible to think the old way
anymore. 

Another great thing about it: once you begin to see it, you suddenly realize
how little you really know, how very much there is to learn.    

=======================================================================

Re: education/help/seddon

John Seddon <101454.573@CompuServe.COM>

16 Oct 96 06:32:22 EDT

I'm looking for help.  I have become involved with educationalists (teachers)
over an issue of ranking.  In short, a student has been demoralised by 'failing'
a test.

I have shown the teachers practical examples of variation affecting performance
in organisations (eg telesales) and they have responded with interest tempered
with understandable reservations ('can see your point there, but this situation
is different' etc).  Behind their thinking lies a raft of assumptions about how
to motivate students and so on, but I want  to start at a place of common
interest.

Does anyone know of published work showing common-cause variation in student
performance on examinations?  I think this would us all enquire further.

Does anyone have a better idea?

Thanks in anticipation.

John Seddon

===========================================================================

Deming in the Classroom/Wintheiser

rickw@individual.eunet.pt (Rick Wintheiser)

Wed, 16 Oct 1996 10:40:19 +0000

I am looking for help in finding how Dr. Deming managed his classroom. In
my university we have constant debates over grading and grading criteria.
In my study of Deming methods I remember him saying, " I can identify the
best and worst but how do I arrange the middle?" He then went on to say
that he gave only passing (I think he said A's) grades. How did he define
his specifications? What was his syllabus like? I fundamentally feel that I
too should shift to this method but I must also defend myself against the
unbelievers. Your collective help will be greatly appreciated.

Thank you.

Rick

Rick Wintheiser            ***********************************************
European University        * Excellence can be attained if you...        *
R.Fernao Mendes Pinto,31   *   Care more than others think is wise;      *
4150 Porto, Portugal       *   Risk more than others think is safe;      *
351-2-610-5728 Voice       *   Dream more than others think is practical;*
351-2-610-2442 Fax         *   Expect more than others think is possible.*
                           ***********************************************
=========================================================================

Re: Levels of Aggregation Problem/Taurman

ILX

Wed, 16 Oct 1996 08:05:26 -0500 (CDT)

At 10:22 AM 10/12/96 -0500, you wrote:
>
>Any thoughts?
>Dr. Jerry R. Goolsby

Sounds to me like you are beginnigt ot understand. The data is only to help
umnderstand the problems. 

If when you start to measure process results and bearing in a machine is bad
then it is part of the system and the chart show a normal variation because
of the bad bearing. Clearly the bad bearing should be a special cause to
management but the chart would show it as a common cause. Yet, if the cahrt
were to show a bearing going bad it would be a special cuase.

Common and special casue depends on where you sit. Deming was trying to
illustrate that managers are responsible for the process people use whan he
spoke of workers being able to take care of special cause and only managers
could fix common cause variation. That is right in theory but in practice it
is hard to tell which is which because so much that is broken in a system
has beocme a part of the system.

When we find that situationit takes us back to the reasons the system that
were not repaired as they deteriorated and we get back to a maangers
derelection of responibilty to put in place systmes that fix things.

Sounds like I ma talking i circles so I usualy avoid beig harsh on the point
that common cuse is managemnt responsibity and the worker can only fix
special cause but can not fix the system.

You are right. Common causae and special cause are dependent on where you
sit and wwhen the data is collected. The idea is valuable wen trying to
track dowwn the cause of variation and imprve the prociess becasue it
spearaes casue of variation into permanet and newly introduced variation.

Back to the bearing. Bad bearings can linger for moinths so one has to look
at those as common or special cause as with many other causes of variation.

Your perspective alwasy depends on where you sit to observe the problem. I I
sit in the corner office a bad bearing may not be so agrivating but if it
impacts my day by causing me trouble all day, it is a very big deal.

Gene  
Gene
Eugene Taurman
interLinx
ilx@execpc.com
http://www.execpc.com/~ilx

==========================================================================

Re: Leavenworth and CLT/Taurman

ILX

Wed, 16 Oct 1996 08:05:29 -0500 (CDT)

At 03:27 PM 10/14/96 -0400, you wrote:
>I had a chance to talk to Leavenworth back in 1991 about his comment on
>the CLT.  He indicated that the real basis for the control chart was
>that it works.  He said that the CLT makes it reasonable for us to do
>things like OC curves, associate probabilities and so forth.
>Later after the 6th edition but before the 7th edition I again had a
>chance to talk Leavenworth about his comment on the CLT.  I pointed out
>to him that what was written in the 6th edition may cause confusion
>about the nature of the control chart.  Leavenworth indicated again that
>the basis for the control chart was not the CLT and that he would make a
>note on the subject.  Hence I believe that is what he attempted to do in
>the 7th edition.
>
>Alson C H Look

You make a very valid point. Well stated. 

I have found the best reason to use control charts is their help in focusing
managers on what is important to the process and helping people see and
understand how well they are doing.

Regardless of any statisitacl theroy the simple line on paper helps us see
change with a higher probability of success than comparing spot data to
speciifcations ro standards. 

Gene  

Eugene Taurman
interLinx
ilx@execpc.com
http://www.execpc.com/~ilx

=======================================================================

Re: W. Edward's Deming Medal Nominations/Taurman

ILX

Wed, 16 Oct 1996 08:16:46 -0500 (CDT)

>
>--------------------------------
>
>Question:  In the debate regarding whether Deming was for or against
awards, the note about Deming "criticizing the philosophical basis" of the
Baldridge award seems significant.  Any idea about what he took objection to 

I do not beleive Deming was againest rewards or awards. His point was very
simple: Incentives and Bonuses do not make peole work harder or better.
There was and is a wide spread beleif that workes need to be made to work.
He was telling us that is not true. Incenitves have failed miserable and
have capped performance.

I believe he alao recognized that people use measures rewards to help figure
out what is important to their own success. So awards are important when
viewed as communication tools. 

He was trying to tell us rewards and incentives are not the answer becasue
they had been used with out understandig human behavior. But what we reward
for is genuinely important to workers understanding what is important to the
organization. 

Gene
Eugene Taurman
interLinx
ilx@execpc.com
http://www.execpc.com/~ilx

========================================================================

Re: Control Chart Usage /Stauffer

Ripstaur@aol.com

Wed, 16 Oct 1996 12:54:30 -0400

Steve Prevette writes:

<< It is still a challenge to see that these hundreds of pieces of paper 
      are not simply wall-paper, and are actually used by line management to 
      make decisions about.  I am currently going through discussions with 
      management about "Simple and Flashy" (that is a quote) performance 
      graphs versus control charts.  I have found the Forrest Gump analogy 
      to be useful in these discussions.   >>

I have the same problem--I believe we work in different parts of the same
organization. At a recent seminar, Don Wheeler said something I have found
useful in dealing with this mindset: "Decoration is not insight." 

I have found, though, that if you can get a leader to really understand what
the control chart is saying, that leader will begin not only to use the
knowledge, but to crave more. I know that's a big IF, but if we didn't
believe it could happen, we wouldn't be doing what we do, right? 

Rip Stauffer
Ripstaur@aol.com

=====================================================================

Re: Levels of Aggregation Problem/Stauffer

Ripstaur@aol.com

Wed, 16 Oct 1996 12:54:36 -0400

 Jerry Goolsby wrote (among other things):

<< Depending on how one defines the boundaries of the system, any 
 improvement effort could be viewed, at least potentially, as 
 "tweaking" a higher system.  I believe it is universally agreed 
 that bottom-up improvement efforts do not yield as fruitful 
 outcomes as do top-down improvement efforts, but isn't it true 
 that every improvement act impacts a system at a higher level 
 of aggregation?   Thus, doesn't the success of any improvement effort 
 depend in large part on its ability to significantly alter the higher 
 system, which makes the effort bottom-up and less than optimal.
  >>

This, to me, is really good stuff, and helps point to the strength of the
SoPK as a system. In his forward to Henry Neave's *The Deming Dimension*,
written in 1990, Dr. Deming said the following:

      " A System of Profound Knowledge consists of four parts:

                   Knowledge about a system;
                   Some knowledge about variation;
                   Some theory of knowledge;
                   Some psychology.

      "The various segments of Profound Knowledge cannot be separated. They
interact with one another. Thus, knowlege of psychology is incomplete without
knowledge of variation.  If psychologists understood variation, they could no
longer participate in continual refinements of instruments for rating people
on a job."

In a later paragraph, he defined a system as, "....a series of functions or
activities (hereafter components) within an organization that work together
toward the aim of the organization. There is in almost any system
interdependencies between the components thereof."

He went on to say, "The aim of a system must be stated by the management
thereof. Without an aim, there is no system. The components of a system must
me managed."

This is why synthesis must be considered before analysis. Without
understanding the system, the aim of the system, and how the components fit
together, optimization of any given component can quickly suboptimize the
system as a whole.

I think Jerry has started to answer the question when he says "Depending on
how one defines the boundaries of the system...."  Without getting too far
afield into system delays or chaos theory, one could make a case that any
action we take might influence the cosmos at some point on the space-time
continuum. If there are definable boundaries to our system, and someone is
managing it, our "tweaking" should have a POSITIVE effect on the system. It
seems, from reading the posting, that NEA is involved. Do they have a role as
"system manager" in this case; are they using Pinella as a small-scale test
("Study" phase of PDSA) for future optimization of a larger system? 

I manage and teach a class for the Department of the Navy called "Systems
Approach to Process Improvement." We try very hard to get our students to
appreciate the systems aspect of the course. Unfortunately, because it is our
only course which teaches SPC, many students come for the tools and ignore
the systems thinking as too deep or philosophical to have any real practical
use. Through a series of practical exercises, those who will listen and learn
the systems side of the course soon learn to appreciate the interaction of
components within a system. They are able to improve the components which
need improvement, thus improving the overall performance of the entire
system. 

Unfortunately, there are those who will not listen; those who have learned
through years of competition, ratings, etc. to make "their" parts of the
system "the best," no matter what. They leave the course frustrated by their
inability to improve (despite "best efforts"), certain that the teams that
did make progress  were just lucky, and even more certain that this approach
just will not work. These are people who insist that there must be a way to
get everyone's performance "above average." They come up to the instructor
during a break following the Red Bead Experiment and say, "OK, what's the
secret? How DO you get 50 white beads?"  They see points above the center
line or below the center line in a control chart and are absolutely sure that
the differences mean something, something should be done, HANG the control
limits, they can SEE the points are different!"

Sorry, I don't mean to get off on a rant, but it seems Pinella is doing some
good things. Jerry says the DOE saw Pinella as a "Special Cause." There's
nothing wrong with that. If your special cause was performance better than
the common performance of the rest of the system, it should be studied to
help improve performance in the rest of the system. It seems the DOE is doing
that.

Sorry for the long posting. I just got back from an extended trip to England
and Italy, and it took me a week to catch up with the backlog of DEN messages
from August, September and October. So, as I heard Ed Baker say once, "I seem
to have answered a lot of questions. I hope some were ones you asked."

Rip Stauffer
Ripstaur@aol.com

======================================================================

Request for randomization story/Stauffer

Ripstaur@aol.com

Wed, 16 Oct 1996 12:54:44 -0400

I recall someone (Myron Tribus?) posting a story about randomization,
sometime last year. They had tried to convince a client to take a sample in
random order. The client wanted to just go A,B,C,D,E. Eventually, the client
was convinced to try randomizing, and the randomization yielded an order of
A,B,C,D,E.
     I have been through and through my archives looking for this story, with
no luck. I either deleted it or it was in some other forum. If it was a DEN
posting, could the originator please send it to me?  Also, could I get
permission to use it in training? I would, of course attribute it to
whomever.

Thanks,

Rip Stauffer
Ripstaur@aol.com

========================================================================

Re: Postings and special cause(s)/Tolman

"Anton O. Tolman, Ph.D."

Wed, 16 Oct 1996 10:42:10 MST

On Friday, October 11, JD Kromkowski said:

> Putting aside any mathematical errors, if a point which falls out of control,
> one may if economical (another key point about cc as a new "economic" tool)
> decide to search for a special cause.  I did -- vacations in August.
>  Confident in my explanation, I decided no further time should be wasted on
> issue.  At that point, the control limits should be recalculated to exclude
> "explained" anommoly.  
> 
> AFTER that, I personally don't think any more calculation of cc should be
> done UNTIL the next special cause arises.  Hence, for future data points, it
> is simply a matter of plotting point.  

As a newcomer to SPC, I would appreciate DEN comments on JD's points 
here, namely that after you understand the special cause and have 
corrected it, you recalculate the chart  the out of 
control point.  I had always understood that you did not need to 
recalculate and did not need to exclude the out of control point 
because the charts are robust and will define out of control points 
even when those points are used in the calculations.  

I agree with JD that you continue to plot the points and do not 
recalculate limits, although I don't recall from Wheeler any specific 
suggestions about how often you recalculate the chart.  I would 
appreciate assistance on this point.  The other question that occurs 
to me is when you implement a change in the process at point X, 
assuming that you have first stabilized the process, do you then 
recalculate the chart from point X onward and compare the mean 
(hopefully moving in the direction you wanted) to the previous mean 
or do you continue to use the old calculations for a while and 
compare the new data points to the "old" limits or the limits of the 
previous process.  At least, this would show you if the "new" points 
were out of control in the old system, for which the special cause 
would be a deliberate shift in the process.  

I am struggling with this because of the level of change occuring in 
my facility; we have a need to be able to quantify how successful our 
efforts are at improving old processes.  Some of these processes are 
stable, and we are trying to shift the whole thing.  So, if I make a 
change in a process, but keep the old control limits and plot new 
points and they fall within the limits, can I conclude that the 
change had no effect?  That we need to search or redesign the 
intervention?  

> ONLY where future data actual signals something (rules for 
grouping, trends
> should be used sparingly in my opinion, but then I'm still a hack) should
> time be wasted on what could amount to a kind of 100% inspection of a stable
> system.
> 
> The danger as I see it is that if one gets involved in constant analysis and
> recalculation we're not doing prediction (i.e. into the future) anymore.

I think this is a good point.  We need to have a balance between 
analysis and planning.  I see the PDSA cycle as helpful here with 
most of the analysis occurring early on in the P and mainly focused 
in the S stages.  This allows some balance and gets the work done.  
________________________________
Anton O. Tolman, Ph.D.
Director, Psychological Services
Director, Quality Management
Wyoming State Hospital
P.O. Box 177
Evanston, WY  82931-0177
Anton@wsh.state.wy.us
(307) 789-3464
---------------    

=========================================================================

Re: Leavenworth and CLT/Kromkowski

Kromkowski@aol.com

Wed, 16 Oct 1996 15:02:39 -0400

Thanks for your other commentary.

It was my understanding that Tchebycheff's Inequality implies that _no matter
what the distribution_, no more that 1/9 of the values, will fall outside 3
sigma.  Where the distribution can be assumed to be approaching normal
 (<--not sure I have this part right) that no more than 3/100 of the values
will fall outside 3 sigma.  If I have this right, this seems to be at the
core of why control charts work.  Perhaps you or others could comment?

JDKromkowski

==========================================================================

Pall bearers and Tom Peters Crazy Times/Sager

Carol Sager

Wed, 16 Oct 1996 15:53:17 -0400

Several people have asked for more information re: my post on this 
thread. The following are chapter titles from  The Tom Peters Seminar: 
Crazy Times Call For Crazy Organizations Vintage Books 1994. My point is 
that for many organizations, the funeral is over and the body 
interred--only they don't know it yet.

BEYOND CHANGE 
Toward The Abandonment Of Everything

BEYOND DECENTRALIZATION
Disorganizing to Unleash Imagination

BEYOND EMPOWERMENT
Turning Every Job Into A Business

BEYOND LOYALTY
Learning To Think 
Like An Independent Contractor

BEYOND DISINTEGRATION
The Corporation As Rolodex

BEYOND REENGINEERING
Creating A Corporate talk Show

BEYOND LEARNING
Creating The Curious Corporation

BEYOND TQM
Toward WOW!

BEYOND CHANGE (REDUX)
Toward Perpetual Revolution

-- 
Carol Sager, Sager Educational Enterprises
http://www.erinet.com/patterwc/CLIIN/
Critical Linkages II Newsletter; 21 Wallis Road,
Chestnut Hill, MA 02167; V.(617)469-9644; Fax(same)-9639

=============================================================================

Re: Use of Surveys/Sager

Carol Sager

Wed, 16 Oct 1996 16:38:09 -0400

> The only personnel survey I have seen was just like this: it was concerned
> with *asking* people if they were satisfied with (a list of things). All
> that could be concluded was that some were more satisfied than others, or
> at any rate more likely to complain. This is just like customers: and, of
> course, personnel are customers of management services.
> 
> If you want to predict action, you should study action, not words. A
> formal survey may or may not be a good way to do this. Since in most
> cases, especially within your own company, you want to know what to
> *try*, in the PDSA cycle, rather than what large-scale decision to take,
> you do not need the right answer, just an interesting theory to try.
> 
> David Kerridge

It might be interesting to start a thread on what the well done survey 
should contain. David, I would like to hear your comments re: the not to 
be missed questions to ask or points to include. 

For my part, I find that it is critical to survey the customer on how 
important the item/product/service etc. is to them as well as how 
satisfied they are with the same. If a customer is 
disatisfied/satisfied it is helpful to know how important that item is 
to them before deploying resources. 

Carol Sager, Sager Educational Enterprises
http://www.erinet.com/patterwc/CLIIN/
Critical Linkages II Newsletter; 21 Wallis Road,
Chestnut Hill, MA 02167; V.(617)469-9644; Fax(same)-9639

[Moderator's note:  As we consider responses to this request, let's try to use
Deming's Production Flow diagram and PDSA as a context.  Thanks   Jim]

=====================================================================

Re: Deployment Flowcharting/Tribus

"Myron. Tribus" <104055.2663@CompuServe.COM>

16 Oct 96 23:10:47 EDT

Here are some sources for software for deployment flow charting.

The mother of all flow charting software for projects is,  in my opinion, a
package called "Team Flow".   The author,  Ron Cordes,  tells me he developed it
after seeing the video tapes on the subject.  (The videos are  available from
Quality and Productivity, Inc. 1081 Westwood Blvd,  Suite 214, Los Angeles,   CA
90024.)  Team flow is sold by CFM, Inc., Sixty the Great Road, P.O. Box 353,
Bedford, MA 01730-0353  e-mail: CFMpro@applelink.apple.com    Team Flow asks you
first to describe the organization chart.  Then you pick out the team members
who are to work on the process.  They are arrayed across the top of the chart.
Then you select some standard symbols to indicate steps in the process and place
them under the names of the people involved in that step.  For each step you
enter a description of the task, start and end dates,  budget,  schedule info,
etc. as well as any quality issues or other references.  As you proceed you mark
off the ends of phases (i.e., planning,  doing, checking, etc.).  As the symbols
are entered the arrows are drawn.  When you are finished the program presents
you with a Gantt chart,  a budget, etc..  I have used it for some engineering
projects.

Team FLOW is not so handy for educational purposes.   For this purpose I have
been using a program called "MacFlow", originally developed for the macintosh
but now a Windows version is available.  The source is Mainstay,  591-A
Constitution Avenue, Camarillo, CA 93012  (805) 484 9428 
 E-mail: 76004.1525@compuserve.com   MacFlow is much less expensive than Team
Flow but has many fewer bells and whistles.   I have used it in interactive
sessions with audiences in which someone from the audience describes how
something gets done and I then draw the flow chart as he speaks.  I do not think
I could do this so well with Team FLOW.

Another program,  apparantly on a par with MacFlow is "Top Down" by Kaetron
Software Corp., 25211 Grogans Mill Rd. Ste 260,  The Woodlands, TX 77380  (713)
298 1500.   I have used it a few times and have gone back to MacFlow only
because I was so much more habituated to it,  not because of any inherent
difference.   Both Top Down and MacFlow are sold by the mail order warehouses
that flood my mailbox with catalogs.


Myron Tribus,  350 Britto Terrace,  Fremont, CA 94539
Ph:510 651 3641  Fax: 510 656 9875
Quoting George Vogler: Your eyes do not illuminate the world.
When you close them,  what happens is not a darkened world.

=======================================================================

Re: Deming Questions # 30/Taurman

ILX

Thu, 17 Oct 1996 09:23:52 -0500 (CDT)

				 DEMING QUESTIONS 

A weekly paper including a set of questions copied in the Deming book "OUT
OF THE CRISIS"   

Chapter 5.  Questions to Help Managers.

30-1.   (No 59 in the Book) - Management by figures

a)     Do you run your company on visible figures alone?

I believe he told us to fix the obvious problems before we even started
charting and measuring the  process results. It is not always necessary to
know how bad it is fix it.

He could also be referring to the interconnection between events. How can
one measure the business lost due to a loose bolt on a car? One can measure
the cost of rework or even the warranty but the real cost is the loss of
integrity and reputation of the business which in loses customers.

Is it necessary to know the cost of disruption in the ship due to
unpredictable raw material quality or should we just fix it.

Is it really necessary to know how much of the sales department's time is
wasted due to poor quality or delivery? Shouldn't  it just be fixed?

Is it really necessary if poor quality is irritating 2% or 10% of your
customers? Or should it just be fixed? Is it necessary to measure customer
annoyance at getting less than expected? Don't they have a right to 100%
good product 100% of the time?  Wasn't that implied when they bought the
product? 

b)     If yes, why?

COMMENTS
He wants us to recognize that people working together are the major part of
making the organization run right Some things need not be quantified but
need to be understood. 
I believe he was talking about setting work standards and using accounting
measures without being aware that they affect attitude and behavior. I think
he had seen companies use accounting figure to drive the wrong and often
poor attitudes and not even be aware of their impact on people.


We do not need to know all the cost or all the numbers to fix a process that
does not work right.
He could say that because he knew if the process up to its' capability it
will cost less and make better quality.

c)     What steps are your management taking to learn the importance of
figures unknown and unknowable?

COMMENTS

He wanted to know if the company was learning about how problems in one area
drive waste or excess cost in another. These are hard to measure but very
real, and very necessary understanding for management to set the right
priorities.

He knew measurements drive attitude and real control is in the measures
selected not in the manager. He wants us to know what the measures really
indicated. In effect he was telling us to use a fish bone diagram on the
measures so we really understood what caused them to change. 

He also knew few had studied the impact on attitude of figures. He knew most
figures measure the wrong results and few studied the causes of variation in
the figures. So he asked how important are the figures you use. 

One of Deming's main thrusts was for managers to know their processes and
how they worked, to know the impact of one part of the system on another.
That means listening to people who have to live in the system. That is not
readily quantified.

Unknowable may have meant too expensive to quantify yet we can know if we
listen to the process by listening to the people in the system. We are not
machines but watery bags of feelings and how well we work depends on being
listen too and understood   


30-2.   (No 60 in the Book) -- Involvement in Standard Organizations

Does your company participate in committees of standardizing bodies? 

COMMENTS

Perhaps this was a very early attempt to bench mark against the best. At
that time industry standards were the most readily  available bench marks.
Many did not and do not even meet those liberal standards. It is also a way
to educate the mangers and a company's resources.

Eugene Taurman
interLinx
ilx@execpc.com
http://www.execpc.com/~ilx

==========================================================================

Re: W. Edward's Deming Medal Nominations/Kerridge

David Kerridge

Thu, 17 Oct 1996 15:59:55 +0100

Eugene Taurman says:

>I do not believe Deming was against rewards or awards. His point was very
>simple: Incentives and Bonuses do not make people work harder or better.
>There was and is a wide spread beleif that workers need to be made to work.
>He was telling us that is not true. Incentives have failed miserable and
>have capped performance.
>
>I believe he also recognized that people use measures rewards to help figure
>out what is important to their own success. So awards are important when
>viewed as communication tools.
>
>He was trying to tell us rewards and incentives are not the answer becasue
>they had been used with out understanding human behavior. But what we reward
>for is genuinely important to workers understanding what is important to the
>organization.
>

This sounds like another of the problems which we often meet: if what
you say is true, rewards may work, but only in a situation that we should
never be in.

If workers don't know what is important to the organisation, whose
fault is that? And if they know, and are not interested in what is
good for the organisation, whose fault again?

David


David Kerridge

British Deming Association Scotland

dfk@rsc.co.uk

=======================================================================

Re: Deming in the Classroom/Latzko

"William J. Latzko"

Thu, 17 Oct 1996 15:04:38 +0000

At 10:40 10/16/96 +0000, you wrote:
>I am looking for help in finding how Dr. Deming managed his classroom. In
>my university we have constant debates over grading and grading criteria.
>In my study of Deming methods I remember him saying, " I can identify the
>best and worst but how do I arrange the middle?" He then went on to say
>that he gave only passing (I think he said A's) grades. How did he define
>his specifications? What was his syllabus like? I fundamentally feel that I
>too should shift to this method but I must also defend myself against the
>unbelievers. Your collective help will be greatly appreciated.
>
Rick,

I took all of Ed Deming's courses at New York University graduate School of
Business. As a requirement for my studies, I had to take one semester of
sample theory with him. Like the man who came to dinner, I stayed to take
all of his courses.

You were required to work. Ed was very patient as to when the work was to be
completed but it had to be done and done well. He required written work
during the course. For the final, he required a paper of our choice approved
by him. 

He read everything with care and annotated the papers. If he thought it was
a particularly good piece if work he wrote a comment suggesting that the
writer should take a Ph. D. Since most of his students were Ph. D.
candidates anyway, this seemed to be a form encouragement. If the paper was
too bad, he asked that it be redone. 

>From time to time, he collected the notes we took in class. Again, he
corrected any errors. If he found that several people misunderstood what he
taught, he reviewed the topic in the next class. 

The material--notes, papers and class work--gave him feedback of what we
learned and areas that he felt needed to be covered once more. In my classes
there were very few times that he had to go back over the work.

If you fulfilled his requirements, you received a grade of "A". Otherwise
you kept working. I know that some of my classmates had constraints that
made it impossible to meet deadlines. He was always patient with them and,
as he told me later, he always received the work and it was always good.

>From the student's perspective, we learned under a master. It was a matter
of pride to us to present the best papers we could produce. His comments
were always helpful. 

Dr. Deming used his texts as his syllabus. While he taught theory, he
illustrated this with examples from his texts and from his current practice.
His plan was to cover the chapters in a logical sequence that developed the
theory and showed how it applied. He developed the basic ideas first so that
we could follow the more intricate concepts when they were presented. 

I hope this gives you some idea of learning that took place under Dr.
Deming's guidance. His statements in his four day seminars reflected exactly
what he did.
Dr. Deming was patient with anyone that really wanted to learn. 

Bill Latzko
---------------------------------------------------------------------------
William J. Latzko
215 - 79th Street
N. Bergen, NJ 07047
Voice:     201-868-5338
Facsimile: 201-868-5338
E-mail:    latzko@worldnet.att.net
------------------------------------------------------------------------------

==========================================================================

Re: Deployment Flowcharting/Kerr

"Kerr, Donald"

Thu, 17 Oct 1996 10:24:00 -0500

Myron wrote:

>Team Flow asks you first to describe the organization chart.  Then you
pick out the >team members who are to work on the process.  They are
arrayed across the top of >the chart. Then you select some standard
symbols to indicate steps in the process >and place them under the
names of the people involved in that step.  For each step >you enter a
description of the task, start and end dates,  budget,  schedule info,
>etc. as well as any quality issues or other references.  As you proceed
you mark
>off the ends of phases (i.e., planning,  doing, checking, etc.).  As the
symbols
>are entered the arrows are drawn.  When you are finished the program presents
>you with a Gantt chart,  a budget, etc..  I have used it for some engineering
>projects.


Here at NASA, I too have used Deployment Flow Charting as an integral
part of engineering project planning. I call it Project Deployment
Wallmapping and I use it as a participatory approach to optimize the
plan.  It is a good tool to transition from the holistic 7 Management
and Planning tools into a reduction to a detailed project baseline
schedule.  Briefly the three major phases of this process are:

1.  WBS Wallmapping - affinity of tasks organized visually by the
integrated product team into an optimal work breakdown structure using 7
MP tools.  Inputs are customer needs, design and operations concepts.

2. Project Deployment Wallmapping- deployment flow chart mapping the
intersection of the WBS with the cross-functional Organizational
Breakdown Structure (OBS).  This is called a Responsibility Assignment
Matrix (RAM), which helps you idenify cost accounts for the project
performance measurement system.

3. Integrated Project Scheduling- takes the RAM, Cost
Accounts,workpackages, and tasks mapped by the team and input them into
the project management software that assigns WBS numbers,
interdependencies, resources, and establishes the project baseline,
which variance is measured against to determine earned value.

There are many project management software tools that enable you to do
this last step as Myron describes.  The best part is that team
participates visually in building the plan holistically upfront, instead
of being handed a reductionist schedule from the project manager at the
start.  It takes more time upfront to involve everyone and do the
mapping, but it is worth it!

We, at the United Space Alliance, are using a tool called P/X, by
Project Software Development, Inc.  It is a Windows application for
Enterprise-Wide Project Management that has a direct interface to an
Oracle database residing on a fileserver.  You establish a Resource
Breakdown Structure and resource availabilities for the entire company. 
All projects share these pools and can tasks can be linked across all
projects, and costs can be rolled up to cost accounts for tracking.

Combining visual, participatory, wallmapping helps establish team
ownership (commitment not compliance) of the plan...P/X helps us manage
the change and capture actuals to continually improve our ability to
predict and reduce variation.

Systems Engineering combined with Profound Knowledge and Quality
Managment tools works!...and you don't have to tell the engineers,
technicians, and operators that they are "doing TQM."  And the best part
is...it is not "Rocket Science!"  It is profoundly simple.

Have a Great Adventure!
Don Kerr

=========================================================================

Resources for Feuerstein materials/Harris

Thomas Harris

17 Oct 96 10:35:59 CDT

From: Tom Harris
Regarding what Myron Tribus is saying about multiple intelligences,
I can verify that what he is saying will change the way one thinks
about the process(es) of education.  I met with Myron last
February and he mentioned MI.  I immediately went home and ordered
one of the books he mentioned above (don't have it here with me).

I've been a teacher for 35 years and, believe me, it changed the
way I now look at my students (and myself, too, for that matter).

There was an interesting piece on TV the other night about a
young lady (classified as a savant) who had genius talent in
the musical field even though she couldn't tie her own shoes or
tell the time.  She could hear a song and immediately play it on
one of several instruments.  My wife (who also read the MI book)
casually remarked, "...her musical intellingence is the only one
which has developed so far."  Interesting comment.

Best of luck.

Tom  Harris

Tom Harris, Ph. D., Prof. Bus. Mgt.
St. Philip's College
1801 M. L. King Blvd.
San Antonio, TX 78231
Voice 210-531-3494
"Continual improvement involving everyone"
...Imai

=====================================================================

Re: WEDI

"Jack R. Stokvis"

Thu, 17 Oct 96 11:38:02 -0400

--Emailer_-1365987201
Content-Type: text/plain; charset="US-ASCII"

I got your note but can't find MIME. So I have reformated by Word 
document to Word for Windows format and hope that you can now open it. My 
mailer only uses Bin Hex, Base 64, UUEncode. I have used the latter. If 
it still does not work, please let me know and I shall try harder.
Jack Stokvis




--Emailer_-1365987201
Content-Type: application/octet-stream; name="ACSP Pap3-1995 / Word 4 wi.SIT";
 x-mac-type="53495444";
 x-mac-creator="53495421"
Content-transfer-encoding: x-uuencode

begin 644 ACSP Pap3-1995 / Word 4 wi.SIT
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M"U]=CQ?[IA@6I>U3PJSI:['V$MY9COWJTS[
M-.O)[VYR,>\ZEM'3>+*11ENG8[)R=.[DPN6MBWI:Y2VCM
MH23N)4M1IM7OV6_NB$O79DY8M;7D^N&NB5NC>-G\DA139O;XW,*4
M6=!SL37?,G.,0J=#T9=.O)'68D:C?%:_+Z?&H?43IJ&Q\8ZF7#I!TU;[[20R
M=V6,3[]@TBZ=>)-=-:=;%W[Y&X+G?:=BIFX:&_2[8_
M[VL#$D";O6;1=:*A&40YP,)9)E3,YC8"@19,1F$8']R;!0J )RU:/F,]8@7P
M8;IGAN9M%2_N&E3M%/#HY8Q@YUX_Q0"QY3BM.OKFX[@V98J!9?8!(QE>RFV.
MVBQW^E&:W$O5?5LR%A7P=0 U8@>@;MH<[0]Z].3U6O@9G)Q22%H'+5OM1"MB
M;
MQ(DV$,W!A L]#]:% \U0[5K/>L 73UC.W(KOLZDK
MN(9"#0>ZR3) X.-,7$!+!S5 S])2]*4[9*06W:IA.I9I]FSBI@>.Z1%3:"IQ
M=+$Y,E ETEA[YUKF+H]&+4K#6+)AZZ;/0/0Z+G5+PRFZV%Z+Z&W;MWY%D39XSJ[8U VH<<:6)=.A&0$V@ F@](2 TY[D=2 U?:!I
M-F%)&?2;F>HJEV-I3G(&-E^!O0'/^U'B\0DCT(+,!L*W
MX2FXM9:4/=JQW.51D>%?BEIN/0E_Z/[G =&]CFOD^AS$/CN=0 !#!)P:!6:!0D$B F 
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MF20N8AKZ/HZ;+8/7"21)K=K:G,N #LVEK@%W)ZM&5B@+FX9:RLD= HQ
MU6\%';5&@M:U(,J#GF)G:^B;-.<"J90.X^A'R'?E6>)64BOGK"
MNI>BTTKXH3W'_#H>912=A$C,TY_% $:SY1DU/E,'2;;
M2)6"V9#!P*HV2AOUO31R:L29I1Y0$Y(JT56=2^!;\8,,EYR'>II.)G95Z]Z&2.T1-#&RBI[-L@
MM_I\B3-!IF !!2B&J\ZTZ3$F&6("^%B# HO1(#1ASB*<&4\"-25;FS;'$F0]
M 9U!@$,R2EP-ZI[):2R]7RTIM:+\>#*!M@,3EXPR<+2,STINDKBF" 23""&0ABVRN.SN%6LW;.3*&P
M#4 ); =X!BQ>[JPW& LD5-2JR;(&A
MA&FK>F.[TIW  40RP"14,I*Z?2E5SYDYO=ZU^>1( _&#HJJQ1"2BGLX8E6^-:-_1.9^.!2""^KN/)><0C>ITX3LD
MT"1;+N37)"Y-\4&ZB>:C3FOL9-3AS:%7:!/_:J4!TOFJ"8.BDD+ 1=0DHC-4
M5CD#HA=W&F) 2\G4-)_KBI@ 7+H16T"ZL5<'MP)M4*]C>L)?!!>)Q\"QJ
M%6%]Z?/@@I+17G:IEZ+*&#F(/\%ZEZM8-=A$>.$*P*2C9&C1P^96D@8SOUW-
M 17)BT 2D7M3AC;J)RE2M@MXBL3I4 [M5G0&E>)"Z4W!RWD5+5E6@$#Q,N"I
M#A.^"E&<437('@#[Q-SWH3DB;7@'/"S?H(!9OW"2/L#7 @2(*]$73&YO:VS 
MM"#Q4S#ENZ19A'9T#@:.\ UD215-G 7[R%!R>N!:./):0?CDKAN%VO!P-X[W8V2>GC)'G;5[&$6^@;.
M5]"4NG'' _DV*Z@DY@+^HW9EN*<^8 #DI
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MJ+?5 OG656B:&RD*25K% .69E333+)92P.Z 0#Q
M7AM;,' RMFJFDJK1![<$&0[Q)$\#8_C$*LA6
M2Z44>. Q*N3!(71EH,A7II]PL1A5:"0";! 7"3?23N%[.\&,XNIL[FI"'"PY
MS9O/ TGF0< ^G" #;^8$\S.R^+"MU1FC3*MFN01-^9Y+5=3^OTVJ5),S%SSG
M6#T8N"R]4"VL0TXD2PF]KG+(L##^G-%Z^V'=+6-[,TLQ> G-@;"P&:%4Y]GLX(S:/8O(E ER228=,>I  
MZC#44&=*_G'OL!@S,,N<"RA1W4+MW535] XF2SP(/D"HY#?WG\(&:&.*40!O
M>=P1M\,/TH@*Q03*GBIB2N<6E=*9TP%3!\CV$ '6-X9390S=X5T3.Q\!5]P(
M["9F8MW89EKS4Y.E V(*:2Y!4)=
MIWTD\7),'2A@,C,:J5))?483?EW#;0VF>&^H:B;)4A;#*)2^$KJ,X[1504K*
MWM@"B'FL;6"NJ"R;G=M\7W2 $A$:84A8<8JF>GH^.)_]6G9U/Z8-0B';2;0W
MQ9,N8([^!@\,,0CT$28OTDKQ'OBQN<#2[.VD&=+^Z%2J7$.P06&<\(6Z&)FU
MX%,<$<*"Z'>[C%N3H39K2B*O+NOF9ECM@OE\R1JGYZGK-MV73X<4RWS$86@T
MAZIZ@()$)>CSHB/(* WT2C(HO9H@\:+&$IH.:)$QYU7;P\\C?7C11F]'
M1T&Y3>"12;6 /5Z[?D0S%:EQCVK$@DPJET1F@W*(>:>=9\6N58EMV:S%63:#
MN5P$X2;(IN"L/E$/!3KC?[#3CEH4":1];?Z<&GD2U"_D;I4)D+U]EOG)]H M
M7R"Y##J0G$_AT..#2:\JU('BDMEM$M[5$#MYHWH_:_JR&R1Q*
MJZ_1XQ0/@MI P&2 H#$T" VY!-T3N,SR#=L6AX7G;%Z0@1ETCV&%6)#&V-@U
M#,;6%M6LZ4L*1+*8&!?B -'/P-R':3,>?+9!)JD@@G;XZ,M(YE)K1K(3=H"_
M"&,?  E4@ HGQ=Q=12&@.GF:Y%S1OMU_M*GFW,IX:%,5EC+( VHKD5O-!D#LA]*YJCN 
M34K =-@+PV=B63TL%YQ59&!0S^%M>FGAL=C8#)-
M#/HPZ[#2D+B(&P62XJT%;&_4X!I^1\4J3,_4A$P+'BH)RG,:B0X.B*XW"5LT
M6T[ BRP"V9JT UC351<9F#H+2RTPLA7P/FY[R%5.=NT9E$=IQKWZ> R#!<,1
MU!>- U +!@VZBFB$IA<4I9L\0H/SG OJM5E/T5N.QYF9L$D=07C=&VFM$&LL
M54"] C&&PXR\A!.1.!PKF4EX'@H16@/M"[>C_5#[5WP_ Y*+_JTJ$C*[X@F$
MR<0.:@)%'A/$((RS"[(:M5+/[ TY%FBIY'Y'X:<
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M 3XQA\2\G1YY7*1RWB*RBVM839W]YJQ8%)A!J"?K!=
MZ"![O%A!A>"V1LP0XB3%7H+&#=)]*M("DH(NN2CN4H-]7M81J8%@J*
M== 2]71OHT6AR ZU*^1J5")["F\), ?Z;48V23:B2^$HOCYCU _QCLK;('.;
M1[A)C,P"2ZW5@P/T8SRV0-3\3RSKDHDILVIDWE#T!=M6IDY#DT9"3PBKP6/,6-=V)4@SD+'(P?/$4/W/+HM))50
ME*KCA  H:+V^PP! LBP,4R(%)3YG[4)UQJ' S)P35/VI9/!RES^#"-QO;/-,
MSS*];U(@J^G"@\E2Y3EJ*1JFLO]6U&K%#9162;YU]3!X0/E;2QF,ZD::7S34HV.V,XVHMY20<"O,!&KK!3"MJP0/(X+EEWA0M
MC:/L+8[OD@\G:Q*UY"]B' ##AXA+TMH&I:H:YJA26 2%2/KK1=.J !O=&# 5
M WR]R2GFEBQ)5^PAR00XC44G2$60$J0*4W+1R$V8<12F972EZG+R[VO&(6*4214DF\1343F%5WI"0Q$3W$?&>**C3:)&XJE,"%<-\D23
M'!HWG@9'IB4Y&DS\&U VW@DO$6:WA%-2X*@-I^950M2&!2\"XSTS8&D(\882
MC)2FR!QSWY-;H)+9JZ3C O#"*%\T[74RSS:YB,A?%;2KP[
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MYD;#PJ30Z:9I(!4VQEOG&=#WM.IC?%)J-1T0RJ-FL\UE,J9SFCLZ+LQX@[** $&Y:B-;$D>(]S3W,XZ0#+JYP?90Q2DBH8VD6#3'9B;AF30CBGQE[@YV
MU6R%[*.N _Z#-CC+248L$,LYS;0KJ39,)5_-+L:<$_[,9;V5VNXI">30E7V$
MVI8B)>"!6"2,KI?58B7($AWW9))RDB3Y:-J<\&I)X(FDMS+EZ#M2,T^HBYJR
M>JU.I\DUR,HOKJHJ*A.10K':G1O,[I/$3$+(!5CM4R;"S.N#\$%XC\F_4'2.
M8&4K,:P"0)#.&=J[E"%65?)\"6
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M:?7,%];P+@Z=)R\$?$!0,%+-AO#-U(4=*O1TYLOS?R6M5 (IJI6#8+(468@9!! I]19"-(PGZ0P$+\SI#D:"I!@MB8XJPR
M"EA!;C]!21SF<*%X"_B$N8X:9A )+4O8&)RY,8""F7N6=RRSD;U2H#C9)R>H0BX91]G.ZSLJD_
MC?C,QNBN+"#@GB+J9<1,./#8=_-%)!'"[IHII)/'53]$C0Y45CC]GL=%Z4$B
M2WF?()#M1\5V3#PR88 *,R68G%:DA%S8(OQYX-249 RQ(/"/N/)20>7=,%
MO@ /3E.";0K.5SX,W0\1OV.K> @ZA-3T@_G857Z4%$/:Q$UYQCD5EE&L>9O0
M;9!ZBT2!59'9?=0S-2<&1PXN%?&OQ
M&X$2MKEQYS$B3SQS\3B1,I/S2:%V&AD1I5 ' $D5\9?8GIO389_<%](>"%_J
M<(GIVDNJO)SB&*H#(?F*34N6_@@+<4W1%XI)CMMH8:45&CPWK+\6JU!(Q1K<
MVB%38!N(TYLF"4 Y]-A[XVI@KM2=)=.9A]P7'MBV/9I8Z9Y-RE<2%U19BN!R
M!HFA-6\)"K)H&=HT#E%IR7 SC$^Z2)79.M 8
M<=C\8+]D955T=9VZ0[T^7]"8P7&FG566"?WR906T$$9:^*).1G$5;]!W$'Z@
M6EA>)6JF )0,L,60GE*V/+4AAQKDC24:"N3Z<]I7R%-\LS.9>048:S968VTC
M8^E'@: E2S-W7@DM*:4H2F[BJ^FFACA4;)FI%#QDVSU0[AB/4220FEZ%SL6Z
M-#98HADQ7(93V0 ?@*0&$O4W28!;4@0>E/B4_)#,UB)F':Z6K;QQ2J*)AN],0TF.JG#IPV_
M&CYENE>5LV<3HD@RE-JZV=A( #.590!59] M\D"%:!
M/!9V:EM+( O;UYPP-Y_EK5E<:1I@VP\H@D,;CA]PWMBR-0>D)IA(2%>R.U>;
M7(V*V!2S7.E":G"8L4Y HITZ:\T2VJ4;PE3;5HD^QG/K8$Q;-$:FE\S-3=#+
M3O48^(?"QVQA#AO[I!4'I;Y^*R@.;1)P);A/7?:9-8'>;.%6@3P]"7'&Z09E
M983("=WJ1X+;>CV)R!'S+/:*/-*S/R 6P3(2)M4FXII&B0^,/C5;853A\LFP
M#BG 8F,\U<.Z[RA^=ACJI6T0@EB4/=L_I(Z2AMKHKMG)6^Z:WUS<@ JQ(V!TL&^U@NV,L!10,#5V&;CH:(-I9R:5O4#,)!J8%6KQ,
M!<7(W6F$Q@7LSR_Y=/J)$O)Y&DF(DPNK,Q<],16TS%RKV1M!0.(4*OR
M9'TT^0:^*T"*:PQD.YEXRH>T+TV8+UNBZ^:SUOY1
M23F.;9I0#X/!+'.674'&!MZK*W#VK"^0H@\ --LNP/31%!F?,>U=Q+)QZ
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M2%>5V23H@!(\8EIO;$3WF[LU^4.8CMET]I)J2N/G(6YLIKC43=*QN,5@GB<@)1#H98O8Q0"A1HB
M9VO8ZF6!CU#Y'7SWQOF(]S#_$:P$ _Q<"X$-RB7C0L@&?KRUT
M5H,VZ.3TRNS%R7(:;3'*C"VE'G85S #,Y2HFA"V"7PE-&" LY#V]Y-3+"U88
M]@%*M$N6IZUZ( "C=TY:VA/#1778" ;Y2+/?70C%]D)1QS=K4@$:!JWST[9,C03QQA*:%4)*XR*B@BK LC*H'2?,FHD),8&"SY1.D20 PJ&\OP
M:.EF(2P!4+M2R#EG$U 030!1VJ?*S8*D=J:L6DXOOE?%ITN;ZU142Q5 "W;G
ME'+;H1.KBR4*M=Z^SQ9\6'!H(W7(<-A^\VH3C<"ZS3-&>TDMX85B9@Q-07=O
MK1%0]B$J>5358C-F( N%?L8[R!HL&N^*!W)E;=U>"2O4=R#7E5MOQ%ZWS)9J(L$Y\!]
M9!Y-=W;+Q82IAEFV5M\H(I]9^1LW#2T>/"HQXZZ
M=%GD8F6SAUQ5<- W[ T-JZ5?+XHP+C1+*C:4ZZ(SL6/2\C"--\BC$5AIG1B(A
MAO4Z^FRK,9H4'VV5-I!7'"F:N\A$,:=A&M''5)JYC 0(M5C8&&H"%7 !K%9K
MRS'WAEWA#AXEF S2.[.5T$YR^YZ%ND%W8]SQ+0!!8[35F4R)(8+"4[)
M=;$2>0N.)W]G]X6L]QGO.4>BJ[CDG1KH+E-Z@33$+WFLFW);>4,Z<:0C;/;-
MPOLLG]%)2>8%Q1-7K TD&WG[M!8Y:Z:OHE@D,31K0Y1*&QC@E$QIN*C%'U@Z
M0X;&S6 [F/6HL<&.Y[H,P;&P%-SRF8W5*F+%!E- I(A@;-,"-^YG)WN/P&Q/
M#P(6ZMU9@;*1.L:MJ6N>V..,2AY(H+]K@5,\2P!HF&($4Q-ZQJ)R\
M.?:&Z.?UV" Z&K\"?1GY_!!%M,($EZU$DE$76P'O&%C0#?\K?#AW:+2,PET*
M#Q1HR)+EH=F$[
M<_$H)1Z.QLBY @NJD[VP<@; '"6&CR1I/R;A45W28G-R=(A.T*NB@/=@3&[_
M%OG>C =P;:4B<7EEDT(Y;(D\K7C'WA9.&P[< IK&AV5T=*]/$?E,U4-(SE:-
MBN0QI$)I)3Y-&_-!IB/3)OU/ZZMX[XT=RK4K99"7>7O(ONV$<\=J9>N>K
MG GU2-#P*#$T!6<<1I(80 C P(\ ,6B1Q01@SX^UJ%K&WGSXHCJ<#!9'00+J
M&8E2IX#4M%\N5N6%/(MA*)Q+FRRE,-GQWFT88"2R,*Q#A*LB3"8501#/*Z+4
M?$3T2%MF).^L5I&HECTQD--R%KL/'6"&^&C%N%.UXYSFG*ZZ:;/!?I2@G1)"/.$#DYR6
M?.4KH*OHKA:43(]LA;::5XH$#CA%D>-(KT2*;O$N@I
M'NY*"-G>C E>%%;?@,I3;1\55KML*DO3G&U)TJW)(S?7Y-MRF*-1L"J(NR9K
MNR]?NK,QCP$#J#E
M4 IY'6?VQV5[)O(O,@LT-
M3-,^A[7$YJL";[HS@Y2/MK^H5AK,"V9H*:)$Z"S &_(7NJ^AX'US/OB<:^NG
M'*J$[77AYU(NE!8,JP*72#G]KJ]#?!Z$17\]'2@65S$%+S.AL!'$AC:(..OI
MDA7X%,TX;\"BP*$.6E7X4JEE9/UI#YS*D+ZY7E].#BCY9EN6'/D&!JY6EKF$
M8-E^CULWY[RJ<(]:SKTQ;Q J)P 1BF%KNKK'";.@]$7&+%H6N/+));5[0 Y7EZ;>4D8%1NO\;9*LI]AJ5'/[%F(!ANY9R.0S9L(
M:8T(OF2;TI0H$2F3TM/"J*Y.:#J+->QX_^Y.3!RJN?
M"&J:E9O.;4PV4D\";C85[\KUY>?MB9GMF/?56K
M2H]V1;T"
M*500+*K-SU+"5U38)&%%< 5ZZ,$(;M:-W:[I@?;+V<(0CYSMB
MA$(%;'<7LG-LP CQP)RUEFH=^S9N#3#)=[E"3Q?^9583;[ICT1&>JUS$V-BB
MSVDE__,&*!GHHL(T;1E_H09W?9*&6 =I'[UCX3:?[E0']UBO;&?.5 /6[/>U
M,V,8;&7$O'JD$\IZT\@,05O^J8>4H\;OE9>U.[O0PR87&DW"I1/CLZI6I"B]
M1S/R('P^EF8W136',-\@2H5V,Z +PH6T4">VO@B)[$AC6.35S/=2
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MMDC5C(O6[O&C.]YMA&/TGVMUQ$H=Z$2#UD1JAK&9=('Z)<;-,0IIA5J#VY^ON_#Y!?YMCH)E[XQ-A-JS/=8X
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M@)%96KIZ9&/75G*F(A 89 N3KPOPRC%,9+59.$[&3*T;5^8UK+(
M&W%9S@P:+#;.^NSE?GTX ,,R(GRWMF3>S>&47BO^/TFVM+S6;K!.FY'Q
M=4_=K3P*&$XL$P5Q!0$EVBDB\IVBC=)5F!-A":CNKTJ[9PF&BM<_^\J+2$<"
M&"ZA%'Z'#3-8EC90*/*-;]"1A\U[C+FTOZ>CQ'0\?V_%>D5=6\HZQ_[YX7I[?'6,B?\CY_G&1#W-^
M^@H1\R*1!MUXO\..<'OE?D5SE_
MZOM$3G&^A_-]G)=\O\@NSBLYW_B#(A_BO.2UW,'YKFG>P3G;$NEP#CB'G,]Y
MG:L-Y[SYJQKEGO\@ASL4#(F_G_/J_?OUK+WQ9OOB/?__Y
MLY][]J_DL_*7\HS\A3PM?RY_)I^1/Y5/RY_(4_+[\JG?D=^6WY)?EX_*K\B'
MY4/RB_*$?%#GZ GC8^-7RS@=GKB6\TEC8V/C5XX)4S1Q_=:]^HWZQ7AM5 ?J
MVO[#+[T77YYRZ<[E_W=[XHCV_PV<;^2\];#(*N?MG.N<;^5\!^=?OD'D&9G(:SBG.%]SLTB;\PV<;^&\Y.N<*T=%WL_Y
M=YR?YWS7[7S'^>N3DKGY/GY&_D
MK^59QMN&^E,ZT'_X:\PTY_D>^5EYOSPBZ_+PQ3G\UCP_,N=Y\IHW)O[_[4=N
MQM_ION#6A+_UM(MO5M\G#NI\_"WG%SA?T!7Y(<[;.=_$>3_G9SGO[(N\D_-/
M,Y'/<%[F1%[&^:6!R#]PMDNP%.?UG"?%[1&[GG/BF
M_(?\N_R;?$V^*N?E*_*"?$G.R1?.BG(!,H _D3^2WY/?E8\#_Q_W\/\1^<@6
M_&_A?TS?%B\X+\'7A[ F\PERD8=8Q<01[?\ASM=R_@SGSW+>QGF"R_GWG%_D
M?!,4_0[.MW#>R7F*\UV<[^9\+^>#SFLXK^5\(>>+.'^2\Z6<^SAOX+R1\P#G%W>(G.-\^U7T"N>=D"GJ
M%0_6OP>^C]_P(YR3G;LZK.7^"
M\SK.-N>[.7^7,W2A9W/^4*T/41OZ,OK/87"F1Z%WO0)LP3RL?#>8 )WGSTW0
M=%['>>VKX60X/_H#X,OS1CK/EW?#-:^$&T+C"4WG2^@YU')^]*#.^C^=.[OQ
M^\S33S_%[[Y/?>QC'PV_#^KO?3_W/GZ_S)@__$L3\\1FA@HB5#-V'H'[&)B$
M,4]\:OSQ'>9ZN?SV?^''P_AP&@VF=8Z>\L,C;>?#Z#C/0*MI'0^ZS.XW@=70
M9;[_SI$N<-\G?F/K]]%?^L!CCP$.]4\3
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M^BG.;]#/S]')?^3\!N>344=^D'.1\X06
M5O.9F#!2IR].7&5D5KVH$]?,R)SNS \,@:5K GG@-=&@.X^]:28J"[F6>F ^
M&NAE;% +OQ$AXBAFZ8F7TF=XYK6 '-QF_:$^_@IG22E#JB/,1VQ5]7WR#OR'
MRO@A+!U:)+(?EC4"Y;'%4\<(OYE D#AFE]4^?9@\'1C*M!(5Z=,MB_OVSH0/
MNV?0&P12.)^I&/&
 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]            

RE: Leavenworth and CLT/Kilkenny
Matt Kilkenny 
Thu, 17 Oct 1996 09:26:00 -0700
[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            

Regarding the comments of Mr. Look....

I had a chance to talk to Leavenworth back in 1991 about his comment on
the CLT.  He indicated that the real basis for the control chart was
that it works.  He said that the CLT makes it reasonable for us to do
things like OC curves, associate probabilities and so forth.
Later after the 6th edition but before the 7th edition I again had a
chance to talk Leavenworth about his comment on the CLT.  I pointed out
to him that what was written in the 6th edition may cause confusion
about the nature of the control chart.  Leavenworth indicated again that
the basis for the control chart was not the CLT and that he would make a
note on the subject.  Hence I believe that is what he attempted to do in
the 7th edition.


I am using the 7th edition of Grant/Levenworth's book in my class at
school.  The chapter on"why the control chart works" talks about the
normal curve (CLT), figuring probabilities for a point falling outside
of the limits, etc. The book is riddled with figuring the probability of
a process shift based off of one data point on a cc, etc.  If
Levenworth's point is as you state, then I don't see it in his 7th
edition.  The book doesn't clarify that control charts aren't based off
of the CLT, it's muddled.  I think it is also a terrible book with OC
curves, sampling plans, etc.   No matter what you think of these tools,
Dr. Deming and Dr. Wheeler tell us they're not statistically valid...the
bottom line is if you think they're statistically valid, you are still
knowingly sending your customer defective product.  Those days are gone.

"Understanding SPC" by Don Wheeler is the best SPC book I've seen on the
market.

matt.kilkenny@readrite.com

=======================================================================




[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            

Leavenworth and CLT/Look
"Look, Alson" 
Thu, 17 Oct 1996 13:47:00 -0400
[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            
I hope you did not misunderstand my comments about Grant Leavenworth's
book.  I was simply stating what he told me.

It is interesting to note that Deming in his famous Out of the Crisis
book recommends avoiding some of the things you mentioned in your email
dated Oct 17, 1996 to me and the Deming list.   Yet  Deming recommends
Ott's book and the ATT book which talk about some of the things you
mentioned.  In fact the ATT book has modified control limits, lot
tolerance sampling plans and OC curves.   I do not believe Deming would
recommend these methods.  The point is as Deming pointed out we should
avoid these things because they will derail self study.  They would
hinder our ability to continually improve our companies.
The control chart is an economic statistical method.  It is based on
some statistical theory and verified by experiment and experience that
it works.  (Shewhart 1931 p277).

Alson C H Look
looka@corning.com
Standard disclaimers 

=============================================================================




[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            

Rebaselining Control Charts/Prevette
Steven_S_Prevette@RL.gov
Thu, 17 Oct 1996 13:55:59 -0700
[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            
     Anton Tolman asked about how to go about recalculating average and 
     control limits following a significant change (such as the August 
     datum in the DEN message count).
     
     I have called this rebaselining, since that is essentially what you 
     are doing -- establishing a new baseline from which to detect further 
     significant changes in the process data (or even a return to previous 
     baselines).
     
     Determining that you have sufficient data to calculate a new baseline 
     can be more of an art than a science.  Statistical theory (Shewhart's) 
     is that you need at least 25 data points prior to declaring that a 
     process is "in control".  But we do not want to, or need to, wait for 
     25 data points prior to establishing a new baseline to look for 
     further statistical changes.
     
     Definitely, after only one month out of control (outside the 3 sigma 
     control limit) you do not know if this is a permanent shift in the 
     data (due to a special cause which must be found), or a short-term 
     shift (still due to a special cause which must be found).  Generally, 
     after 7 months I have found it practical to calculate a new average 
     and control limits (this fits well with the rule of thumb that 7 
     months in a row one side of the average is significant).  Usually I 
     circle the 7 data points on the 7th month, then on the 8th month 
     calculate the new average and control limits.
     
     So there may be a period of time for which you need to annotate on the 
     graph that a significant change has occurred, but you do not have 
     enough data stable about a new mean to calculate a new baseline.
     
     Note that the prior time period used (say Jan 96 - Aug 96) should not 
     overlap with the new time period.  For example, showing a graph with a 
     Jan 96 - Aug 96 baseline, followed by a Jul 96 - Dec 96 baseline would 
     be incorrect.  But a gap between the two baselines is allowable (Feb 
     93 - Apr 95 baseline followed by a Jul 95 - Jan 96 baseline).  Note 
     that the May and June 1995 data would be plotted against the Feb 93 - 
     Apr 95 average and control limits.  
     
     I should point out a common error is to keep recalculating the average 
     and control limits as new data is obtained.  This should not be done, 
     the average and control limits should be based upon a fixed time 
     interval, until a statistically significant change occurs.
     
     If you have identified the special cause (and when it occurred), that 
     should definitely be used for the split in the two baselines.  Other 
     than that, you are forced to determine a split by looking at the data. 
      There are dangers to splitting based on the data.  One rule they 
     teach statisticians is to never test for statistically significant 
     changes based upon patterns seen in the data after you look at it.  
     This can lead to falsely declaring random fluctuations as significant 
     (there is a agriculture parable that illustrates this).
     
     I do base this writing on my experience so far, and statistical 
     training in SPC at the Naval Postgraduate School, which was based upon 
     Acheson Duncan's >Quality Control and Industrial Statistics<.  I hope 
     what I have written may be useful.

     - Steve Prevette
       Fluor-Daniel Hanford
       steven_s_prevette@rl.gov
     
    ========================================================================
 




[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            

Posting & Special Cause(s)/Crow
James Robert Crow 
Thu, 17 Oct 96 17:47:10 EDT
[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]             
When you have identified and taken action which you believe will remove a
special cause you are at the Study point of PDSA.  Did you action achieve
the desired result?  If so you Act and make this a part of your established
policy.

You now have a stable system, which means the variation is being caused by
the system/process itself.  If you want to continue to improve the system,
(remove variation) we must do this by improving the system.  As we make
changes to the system we should continue to study the impact each change has
on the system.  Did this change result in a more stable (predictable ) system.

Even though a system is stable it may not be as predictable as we would
like.  One of Deming's concepts is that the job of management is prediction.
Obviously the more stable a system the more predictable.  A company keeping
a control chart showing that it is able to consistently deliver its product
to the customer 98.5% of the time can say with assurance that the order will
arrive on the specified date, where as a company with a 70% on time delivery
record cannot say this with the same level of confidence.  Both of these
systems may be stable, however the control limits for the company delivering
its product on time only 70% of the time will be much wider.

As to when you should start a new chart I am of no help.  Several years ago
while working with a company we did control charts on their ability to
deliver replacement parts to their customers, and how long it then took them
to invoice the customer.  The control charts showed that both systems were
out of control, the range could be anywhere from one day to literilly
months.  The average time from when the order was received until shipped was
over 6 days.  The average time to invoice the customer was even longer - 10
days.  Management was alarmed.  We can't do business like this!

We defined each system, made two changes and studied the result.  The
original data was for 5 months.  The changes were implemented around the
10th of July.  When we looked at the data for July it showed wide variation
for the first ten days then almost no variation at all.  Time to ship had
dropped to one day.  Time to invoice had dropped to 3.5 days.  When
calculating the control charts for July I only used July's figures.  They
showed that for the first ten days of the month the system was still out of
control, but became very stable after the tenth.  The control limits had
been substantually reduced also.

Having achieved the desired result we adopted the new procedures.  This is
our order fill process.  This is our invoice process.  I would say that once
you have adopted the new process you then need to do new control charts to
monitor the new process.
Comments?
Robert Crow,            The Crow Group
jr1crow@mindspring.com, Continual Improvement
Voice (770) 461-1177,   Optimization of the System
FAX   (770) 461-1177,   Focusing Organizational Energy
250 Thornton Drive,	Motivating Work Environments
Fayetteville, GA 30214

========================================================================




[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]             

Levels of Aggregation Problem/Crow
James Robert Crow 
Thu, 17 Oct 96 18:05:25 EDT
 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]             
       In discussing the boundaries of a system, one of Deming's points is
that you should work to improve the larger system.  In your case by
improving you system you are also improving the larger system.

Another point Deming makes is that the larger the system the bigger the
potential results, but the harder to manage the change effort.  The
transformation must start with you.  You begin to change your system, which
is working within and in some cases restricted by larger systems.  By
improving your system you will hopefully get the attention of the larger
system which will say what are you doing to get these results?  Are there
things we can use to improve the larger system?  The down side of this is
that the larger system will feel threatned and say stop doing that, we don't
care what's happening we don't do things that way here.

Fortunately for you the larger system seems to be responding favorably.  To
change systems you have to rock the boat.  You must be willing to take risks
with the possibility that you may be wrong.  That's why you shouldn't begin
to change systems without first having a good understanding of the system of
profound knowledge.
Robert Crow,            The Crow Group
jr1crow@mindspring.com, Continual Improvement
Voice (770) 461-1177,   Optimization of the System
FAX   (770) 461-1177,   Focusing Organizational Energy
250 Thornton Drive,	Motivating Work Environments
Fayetteville, GA 30214

=========================================================================




 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]             

Use of surveys/prevette/Crow
James Robert Crow 
Thu, 17 Oct 96 18:06:29 EDT
[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]             
I have used employee morale surveys, but this was before I know Deming existed.

Employee morale surveys can be used to determine is a problem exists and
help you pinpoint what the problem is.  If you have indicators that you have
a morale problem, ie. high turnover, high absenteeism, lots of accidents,
worker comp rates going through the roof. You can as management develop a
theoy that you have a morale problem, and additional theories as to the root
causes of these problems.  You must then begin to investigate these theories.

One thing you can do in implement an exit interview policy to determine why
people are leaving your company.  This in usually done by HR within the
company but can be done by an outside firm.  A friend of mind tells me that
you will get the most candid results by using an outside firm and contacting
the person approximately 90 days after they have left your company.

But you can't wait that long.  You need information now.  This is a crisis
situation.  One of the most critical things about doing an internal employee
morale survey is the commitment by management of study and respont to the
results of the survey.  Doing the survey will raise the employees
expectations.  They will expect change to take place.  If management ignores
the survey results, the result will be that you situation will be
substantially worse.

You need a pretty good system for administrating the survey.  You must
protect people's privacy.  They must be assured that their answers cannot be
traced back to them, while at the same time being able to know that this
department/shift, section is responding differently from the company as a whole.

You must need to know how to interpertthe results.  A question may be:  I
know who my supervisor is?  A substantial negative rating here could mean
that these people are receiving conflicting orders from different
supervisors.  It may mean that the manager above the supervisor frequently
by passes the supervisor and goes directly to the people when he wants
something done.  In other words a substantial negative result to this answer
may call for further investigation to pin point the cause before corrective
action can be taken.  After the corrective action has been taken it in
important to follow up to see if your action has been effective.

I have several examples of employee morale surveys I have used.  Some time
in the future I will put them on the net.

Some company's as a matter of policy do employee morale surveys on a
scheduled basis.  I once was with a company that had each of their
distributions centers surveyed ever 18 months by an outside firm.  For a
long time nothing was done, except that headquarters would review the report
and it would be forwarded to the DC manager.  He could choose to do what he
wanted with the report.  Fortunately most of them took the process seriously
and responded to the report.

Shortly before I left the company we modified this approach.  When the
report was received at headquarters, we (HW HR) would review the report and
come to consensus at to what the report was saying.  I would then sent a
copy to the DC manager, and follow this up with a personel visit.  The DC
manager and I would spend a day going over the report and discussing what we
saw as positive and negagive aspects of the report.  We would then break the
report down by department, and have the department managers come in
individually.  They were told to study the report with their management team
and come back the next day with their analysis of the problem areas, and
what they were going to do the correct the problem.

The DC manager and I would listen to each department managers report to see
is we agreed with their analysis of the problem areas.  We would point out
areas in which we saw problems that the department manager did not see.  We
would then listen to their plan for correcting the problem, and offer
additional suggestions when required. This process provided us with much
better response to the results of the morale survey.

The most critical thing about using an employee morale survey is that
management have a commitment to listen to and respond to the results.
Robert Crow,            The Crow Group
jr1crow@mindspring.com, Continual Improvement
Voice (770) 461-1177,   Optimization of the System
FAX   (770) 461-1177,   Focusing Organizational Energy
250 Thornton Drive,	Motivating Work Environments
Fayetteville, GA 30214

=======================================================================



[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]             

Level at Which to Begin Implementation/Eubanks
Lori Eubanks 
Thu, 17 Oct 96 19:13 EST
[Next of thread]             
 [Author List]                        
[First of thread]            
[First by author]            

I work at a major law firm where we have retained a consultant to assist
us in implementing Dr. Deming's philosophy.  While I have read and been
advised by an expert in the area that this level of profound change will
not work in an organization unless senior management is involved, visibly
supports the philosophy, and advocates the changes that are occurring,
our consultant (when asked by a member of the administrative staff why he
was beginning at the staff level and not at the attorney level, where
many of the staff perceive the problems to lie) advised us that law firms
who have started at the top have accomplished nothing.  He explained that
by starting at the staff level, the attorneys will see the
results/successes and then ask to be incorporated into the new system so
they can achieve the same results for themselves.  (Remember trickle-down
economics?  This idea almost seems like trickle-up TQM.)

Does anyone have any thoughts on the above scenario?

Thank you for any insights you can provide.

Lori Eubanks
leubanks@net1.net

===============================================================================



[Next of thread]             
 [Author List]                        
[First of thread]            
[First by author]            

Aggregation & Deming's Class/Kline
Bob Kline 
Thu, 17 Oct 1996 21:03:52 -0400 (EDT)
 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]            
Concerning the question about aggregation and the linkages of systems and
subsystems, I think we have to watch out for analysis paralysis.  Set the
boundaries, the context, get what you think is the proper data from the
environment of that system and go for it.  If found in a free-flowing mood,
you could find yourself talking about galaxies and universes.  For me, to
make sense and to be practical, I try to keep the context at an arm's length
of control and influence.  Not to trivalize the topic, but I too have had
such conversations and some seem to go much better after a few of my
favorite, frosty beverage.  And part of the question that is being asked in
that discussion is:  How do we know if we are truly working towards
optimization?  Hmmm.

Concerning the question about how did Deming run his classes, I think Bill
Scherkenbach (close in spelling?) was a student of Deming's and he might be
able to pass on some insight.  If you are able to do that, I think all of us
would be interested in hearing about it.

Bob Kline, DON TQL, ISTJ
bobkline@erols.com  
Cheers, Bob.  (bobkline@erols.com; BobKl@aol.com)

=========================================================================




 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]            

Deming & Baldrige/Monroe
PHILHOOVER@aol.com
Thu, 17 Oct 1996 22:05:27 -0400
[Next of thread]             
[Next by author]             
[First of thread]             
[First by author]            
Lloyd Dobyns and Clare Crawford-Mason's book, "Thinking About Quality" has a
chapter titled 'Deming and Baldrige'.  There are a number of quotes they use
from a discussion Dr. Deming had with Curt Reimann, the "Father" of the
Baldrige criteria.

It is clear in that chapter that Dr. Deming did not endorse the Baldrige
award.

Sadly I think, the authors parted with Deming and concluded that on balance,
the Baldrige award brings about more good than harm.

Phil Monroe
PhilHoover@aol.com


[Moderator's Note:  I forwarded this message to Clare for comments.]
==========================================================================




[Next of thread]             
[Next by author]             
[First of thread]             
[First by author]            

Deming and Education/Kerridge
David Kerridge 
Fri, 18 Oct 1996 09:34:41 +0100
[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            
Dear Bill,

        Thanks for your comments about Deming and learning. I did
not experience his teaching in the classroom, as you did, but I
was taught by him, as everyone about him was. He made me think in
a way that unlocked mental blockages that had held me up for years.
For example, I had been struggling with basic problems in the
theory of probability - how probability mathematics links to
reality - for over thirty years when I met him. One question from
him "Why does probability theory work?" simple as it sounds,
crystallised the problem exactly: within six months I was back
on the trail that had eluded me. All I need is another thirty
years to finish the job.

    On another occasion, I told him about something I had worked out.
He made me feel as if I had won the Nobel Prize. Then a long time
later I discovered that he had published it in the 1930's. To him,
what mattered was that we learned to think for ourselves.

David

David Kerridge

British Deming Association Scotland

dfk@rsc.co.uk

==================================================================




[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            

Re: Optimising the whole system/Kerridge
David Kerridge 
Fri, 18 Oct 1996 10:00:19 +0100
[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            
Bob Kline says:

>Concerning the question about aggregation and the linkages of systems and
>subsystems, I think we have to watch out for analysis paralysis.  Set the
<>
And part of the question that is being asked in
>that discussion is:  How do we know if we are truly working towards
>optimization?  Hmmm.
>

I think the phrase "analysis paralysis" is very acute. Why do you suppose
that Dr Deming was so dismissive about "Quality tools"? I remember him
saying "You can learn the lot in 15 minutes"

Plainly this is not true if by it you mean having a detailed knowledge
of all the tools you can use to analyse systems. But perhaps (my guess)
is that he meant that analysis of the particular system is not something
to spend too much time on. Deal with problems that are easy to see, but
mainly rely on general principles than apply to all systems: such as those
expressed in the 14 points and the SoPK.

For example, reduce chaos (SPC, Operational definitions, standardisation),
simplify (fewer suppliers), eliminate conflicts of interest (bonuses),
improve information flow (Break down barriers)and so on. By the time
you have done all this the remaining problems in the system will be
very obvious without much detailed analysis.

I think by training and culture we just love detail, and are suspicious
of principles. But with the principles we know we are moving towards
optimisation, and with analysis of detail we are often guessing - have
we seen all the things that matter? And have we drawn the boundaries
correctly? Principles eliminate guesswork.



David Kerridge

British Deming Association Scotland

dfk@rsc.co.uk

=======================================================================





[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            

Leavenworth and CLT/Look
"Look, Alson" 
Fri, 18 Oct 1996 09:11:00 -0400
[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            
To quote Shewhart  (1931 p 266) concerning Tchebycheff's inequality:

	we should expect to find more than 89 per  cent of the total area for
each distribution within the limits.  In fact, no matter what
distribution we might construct with average xbar and standard deviation
sigma, we would find that more than 89 per cent of the area would fall
within the dotted limits.

There are graphs of different distributions on p266.

I believe Shewhart used the Tchebycheff as a lower bound (p 277) since
we in general do not know the distribution in the real world with
sufficient  detail.

Shewhart (1939 p 36)  makes the statement:

	The control limits as most often used in my own work have been set so
that _after_ (My emphasis.)  a state of statistical control has been
reached, one will look for assignable causes when they are not present
not more than _approximately_ (My emphasis.)   three times in 1000
subsamples, when the distribution of the statistic used in the criterion
is normal.

Shewhart apparently links statistical control with _approximately_
3/1000 subgroups and when the distribution is normal.

Grant Leavenworth (1980 p 78)  states:

	It is not strictly accurate to say that in the long run 3-sigma limits
on an xbar chart will sow points out of control only 27 times in 10000
(that is, 0.27% of the time) provided the universe is really unchanged.
This would be strictly true only if the distribution of the xbar values
were _exactly_ (My emphasis.) normal and the control limits were based
on known values of xbar and sigma.  Actually although the distribution
of xbar values is roughly normal, it is not exactly so unless the
universe is normal; the 3-sigma limits are necessarily calculated from
the observed data rather than from parameters of the universe.  Hence
3-sigma limits may give false indications of lack of control somewhat
oftener than is indicated by the normal curve.
	Nevertheless, such false indications of lack of control will be
infrequent.

Shewhart settled on 3-sigma because it balanced the two economic mistakes
that Deming (1994, p. 99) talks about.

Three sigma limits were chosen by Shewhart, Grant, Deming, and others
because as Grant Leavenworth (1980 p 110) nicely summarized

	The real basis for the use of 3-sigma limits on control charts for
variables in industrial quality control is experience that, when closer
limits, such as 2-sigma, are used, the control chart often gives
indication of assignable causes of variation that simply cannot be found
whereas, when 3-sigma limits are used the points fall out of control, a
diligent search will usually disclose the assignable causes of
variation.


Alson C H Look
ejla29b@prodigy.com
Standard disclaimers

===========================================================================



[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            

Re: Posting & Special Cause(s)/L'Heureux
"Howard A. L'Heureux" 
Fri, 18 Oct 1996 06:17:29 -0700
 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]            
James Robert Crow said:

Even though a system is stable it may not be as predictable as we would
like.  One of Deming's concepts is that the job of management is prediction.
Obviously the more stable a system the more predictable.

I find this to be insightful.

I'd like to add that more stability will yield process capability.  A capable
process, is one which is stable and meets customer requirements (customer key
quality characteristics).  Capability is what keeps customers coming back for
more; but, it can only come from continuously improving processes.

Howard A. L'Heureux
Organizational Systems Alignment

=====================================================================




 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]            

Re:  Concerning revising control limits./Look
"Look, Alson" 
Fri, 18 Oct 1996 09:28:00 -0400
 [Topic List]                          
[Next by author]            
[First of thread]            
[First by author]            
A good place to look at is Wheeler and Chambers, Understanding SPC pp
226-230.  When applying these points try to think in terms of
principles.

Alson C H Look
ejla29b@prodigy.com
Standard disclaimers

===================================================================




 [Topic List]                          
[Next by author]            
[First of thread]            
[First by author]            

Re: education/help/Harris
Thomas Harris 
18 Oct 96 09:15:24 CDT
[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            
From: Tom Harris

Regarding the teachers and the students being motivated, maybe you
should start with the teachers by informing them that they cannot
motivate their students--the students motivate themselves only
after the teachers impart a goal which the students choose to
try to achieve. In addition, if the teachers you're working with
are in a system which doesn't have competency-based education (or
outcomes-based education) as the method of teaching/learning,
there may not be much you can do to change that system.  In the
systems which don't use competency-based education, learning is
variable because the time to learn is fixed; therefore, grades
are used to differentiate between those who learn the most and
those who learn less.  We all learn at different rates, but the
systems which are time-based do not take that into account.  In
such a system, the teachers probably think they're in control
of the learning--'t ain't so.

Suggest you read *Orchestrating Learning with Quality* by David
Langford and Barbara McCleary.  Be forewarned, it's a major leap
forward and doesn't always set well with those stuck in the old
mode of doing things...or as the old Chinese proverb states, *it's
very dangerous to try to leap a chasm in TWO bounds.*

BTW, where are you located?  Maybe we can put you on to someone
who knows how to do some of these things.

Tom Harris

Tom Harris, Ph. D., Prof. Bus. Mgt.
St. Philip's College
1801 M. L. King Blvd.
San Antonio, TX 78231
Voice 210-531-3494
"Continual improvement involving everyone"
...Imai

=====================================================================



[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            

 Deming in the classroom/Harris
Thomas Harris 
18 Oct 96 09:24:51 CDT
[Next of thread]             
[Next by author]            
[First of thread]            
[First by author]            
From: Tom Harris

It IS risky to take on such a move if you're the
first one in your school to try it.  I've been doing it for 2-3 years
now and I've been accused of being a "curve-raiser" and all the
other names used to describe someone who is "different."

However, once you concentrate on your students and really push them
to achieve, you won't have any trouble justifying the grades they
receive.  I'll put my students up against any other group of students
studying the same area and I'll guarantee that my students can win
almost any kind of competition.  They really do a super job when
we "cut them loose" and let them go at their own speed.  They exceed
my expectations by a long shot.

Good luck.

Tom Harris

Tom Harris, Ph. D., Prof. Bus. Mgt.
St. Philip's College
1801 M. L. King Blvd.
San Antonio, TX 78231
Voice 210-531-3494
"Continual improvement involving everyone"
...Imai

===================================================================




[Next of thread]             
[Next by author]            
[First of thread]            
[First by author]            

Award impact/Crow 
James Robert Crow 
Fri, 18 Oct 1996 11:00:18 -0400 (EDT)
 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]             
>This sounds like another of the problems which we often meet: if what
>you say is true, rewards may work, but only in a situation that we should
>never be in.
>
>If workers don't know what is important to the organisation, whose
>fault is that? And if they know, and are not interested in what is
>good for the organisation, whose fault again?
>
>David

I think David is correct here.  The problem with awards is that you create
a win/lose situation.  Yes we have a winner, mabe an individual or team, but
you also have losers, and the losers will greatly out number the winners.
You also have suboptimization in that there is usually a very involved
process to select the "winners".  This is time taken away from improving
processes, and meeting the customers needs.  Rather than having awards for
individuals or teams it could be more effective to concentrate on improving
the total system, and make it into a win/win situation for everyone.

Robert Crow,            The Crow Group
jr1crow@mindspring.com, Continual Improvement
Voice (770) 461-1177,   Optimization of the System
FAX   (770) 461-1177,   Focusing Organizational Energy
250 Thornton Drive,	Motivating Work Environments
Fayetteville, GA 30214

====================================================================





 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]             

Re: Deming in the Classroom/Crow
James Robert Crow 
Fri, 18 Oct 1996 11:05:13 -0400 (EDT)
 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]             
Rick, I am sure that this will not be much help except that I remember when
in college the professor of a behavioral phychology course announced the
first night of class that everyone would make an A.  That is you would make
an A provided you attended the course and stayed with the course for the
full quarter.  There were no tests, or research papers.  We spent the whole
quarter getting to know ourselves.  In fact we did not get enough during the
quarter.  The class met on a Saturday after the quarter was over to do some
more experiencial exercises.

Knowing up front that you did not have the pressure of grades allowed you to
relax and participate fully in the class.  I consider this class one of the
most beneficial that I took in college.  This class changed my whole
perspective on life, and was a valuable learning experience.

I had one other class while at Georgia State that was taught in a simular
manner.  The class was mgt. 350, or your basic management class. The
professor had us give him a one paragraph resume'.  From this he assigned us
to teams, then assigned us a section of the book to teach the rest of the
class.  He also assigned a team leader in each case.  The professor's role
then changed to that of a consultant to the various teams.  We then met
during the normal class period to plan how we would teach our subject to the
class.  We soom found ourselves having to schedule additional meetings to
prepare for our presentation.

The method we chose to teach our subject was up to us.  Our group had
planning.  We developed a case study using a soft drink company.  The
company made a cola, but was considering introducing a lemon-lime.  The
group then examined the in's and out's of the decision, and developed an
implementation plan.

We developed the case study, put in all of the material regarding decision
making and planning and then implemented our plan.  We had one week to teach
the rest of the class how to plan.  Since this was an evening course, we had
three 1.5 hour class periods.  We were successful in that each of the other
teams were able to take the material and apply it to the case study.  Each
developed a plan which they had to defend to the other teams in the final
session.

The Professor's name was Don Jewell.  Don did some of the pioneering work in
self managed teams, and as far back as the late sixties had some plants
operating with virtually no management.  The work teams ran the plant and
the plant manager functioned as a consultant to the group.  Don is still
actively consultanting and is very familuar with Dr. Deming's work.  He
recently completed working with a company in South Carolina in which they
eliminated budgets, (a source of suboptimization) and managed the company
through the key cross organizational processes.

It was from Don that I learned about boundary management.  The basic idea is
that your employees should run the operation and management should
concentrate on the larger system.  What is happening in our industry, that
could threaten our future?
How will technology enable us to be more productive?  What is happening
outside of our industry that could enable us to be more effective in meeting
our customers needs.  This ties in with Deming's concept of the boundaries
of the system and working to improve the larger system.

Robert Crow,            The Crow Group
jr1crow@mindspring.com, Continual Improvement
Voice (770) 461-1177,   Optimization of the System
FAX   (770) 461-1177,   Focusing Organizational Energy
250 Thornton Drive,	Motivating Work Environments
Fayetteville, GA 30214

===========================================================================




 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]             

Re: Deming & Baldrige/Tribus
"Myron. Tribus" <104055.2663@CompuServe.COM>
18 Oct 96 12:27:11 EDT
[Next of thread]             
[Next by author]             
[First of thread]             
[First by author]             
On balance,  I too,  Dr. Deming not withstanding,  believe that the existence of
the Baldrige prize has done more good than harm.   

What are the arguments on the other side,  other than Dr. Deming's antipathy
towards the Baldrige prize?

Myron Tribus,  350 Britto Terrace,  Fremont, CA 94539
Ph:510 651 3641  Fax: 510 656 9875
Quoting George Vogler: Your eyes do not illuminate the world.
When you close them,  what happens is not a darkened world.

===================================================================




[Next of thread]             
[Next by author]             
[First of thread]             
[First by author]             

Re: quality and the law/Tribus
"Myron. Tribus" <104055.2663@CompuServe.COM>
18 Oct 96 12:27:14 EDT
 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]             
You can begin anywhere.  Ultimately,  to obtain the best results,  you will need
leadership from the top.  (See my postings on 'Changing the BOS')  But,  you can
begin anywhere and if you are to begin at all,  you have to start wherever you
can.   Starting and finishing are different stages.

If you wish to connect with another group working on quality and the law,  write
to      
Sam Welch      rblwelch@xavier.xu.edu
who is now involved with a bevy (crowd, phalanx, pride, group ??) of lawyers
involved in the same quest.

In addition,  about three years ago there was an ABA team working on quality and
the law.   They produced a videotape report summarizing the experiences of ten
law offices which attempted to use quality methods.  

They found good success in such things as better operation of the mail room and
the copy machines,  but not much in changing how lawyers deal with clients.  The
biggest barrier was the habit of rating attorneys on 'billable time' which meant
that each attorney did his best to charge his time to someone.

When it came to applying quality principles to the law court (i.e., scheduling,
etc.) they gave up.

One anecdote that sticks with me:  One law firm,  at the conclusion of a court
case,  give a questionnaire to the opposing time,  asking for a critique of
their approach:  Had they been properly prepared?  Had they pursued a good
strategy?  etc., etc.   Needless to say,  the other side was amazed and
perplexed.

Myron Tribus,  350 Britto Terrace,  Fremont, CA 94539
Ph:510 651 3641  Fax: 510 656 9875
Quoting George Vogler: Your eyes do not illuminate the world.
When you close them,  what happens is not a darkened world.

====================================================================



 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]             

Re: Deming and Education/Constantine
John Constantine 
Fri, 18 Oct 1996 10:07:36 -0700
 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]            
David Kerridge wrote:
> 
>     On another occasion, I told him about something I had worked out.
> He made me feel as if I had won the Nobel Prize. Then a long time
> later I discovered that he had published it in the 1930's. To him,
> what mattered was that we learned to think for ourselves.
> 

When I read this I thought to myself - "what a beautiful way to express 
what all our education should be about, as well as our educators." Thank 
you very much, David.

If you know x, you may predict with y with some assurance. If you can't, 
you can't. That's why "knowing" x is so vitally important. 
-- 

Regards, 
John Constantine
rainbird@trail.com
Rainbird Management Consulting
PO Box 23554
Santa Fe, NM 87502
http://www.trail.com/~rainbird
"Dealing in Essentials"

=======================================================================




 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]            

Causing attitude/Taurman
ILX 
Fri, 18 Oct 1996 16:17:46 -0500 (CDT)
[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            
Taurman response

There is no doubt that rewards and bonuses cause behavior. Most often
behavior not in the best interest of the organization, but always in the
best interest of individual. 

Read Alfie Kohn's book on rewards. It is full of examples, though he does
not point them out, of poor behavior due to bonus and incentives. He
advocates no rewards.  

Kohn is very much against reward systems. There is a big problem whith his
conclusion. Every organization past and future has a recognition or
consequence system. Management's only alternative is to know the system and
then manage it. 

Every day people say such things as "please", "thank you" and "that was
stupid". These though unintentional are are consequence systems. They along
with the more organizational consequnces such as attaboys from a boss cause
attitude. They tell the employee waht behavior it takes to avoid risk to get
promoted to be recognized or not depending on the individual's needs. Reward
systems exist and must be managed in detail because they cause attitude.


Gene 

Eugene Taurman
interLinx
ilx@execpc.com
http://www.execpc.com/~ilx

============================================================




[Next of thread]             
[Next by author]             
[First of thread]            
[First by author]            

Re: Optimising the whole system/Daszko
Mdaszko@aol.com
Fri, 18 Oct 1996 19:31:55 -0400
 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]            
In a 2-day workshop led by Dr. Russell Ackoff  this week I learned more about
systems thinking and the "analysis paralysis" we can get ourselves into.  Dr.
Ackoff's books give more insight into thinking about system optimization vs.
breaking systems into parts and trying to analyze them incorrectly.

Marcia Daszko

=======================================================================




 [Topic List]                          
[Next by author]             
[First of thread]            
[First by author]            

Re: Level at Which to Begin Implementation/Delph
"Steve Delph (PAD)" 
Fri, 18 Oct 1996 22:10:17 -0400 (EDT)
[Next of thread]             
[Next by author]            
[First of thread]            
[First by author]            
On Thu, 17 Oct 1996, Lori Eubanks wrote:

> I work at a major law firm where we have retained a consultant to assist
> us in implementing Dr. Deming's philosophy.  While I have read and been

> Thank you for any insights you can provide.

Lori,
I don't know how much insight I can give you but it seems to me that
the consultant is going about this project in the wrong way.  I dont't
have a large working knowledge of the Deming Philosophy but if you look at
it Deming advocates changing the entire organizational system, not just
portions of the system.

I work in a law enforcement agency that, as does your organization,
provide a service to its "customers."  The Deming philosophy advocates
that the system provides "quality service" to all its customers, both
internal and external.  It seems to me that, based on my limited
knbowledge of this philosophy, the consultant is missing the boat and
needs to look at the overall situaion.  I believe that Deming himself
would demand to work only with the chairman or top person(s) within an
organization.  If he could not work with them in an effort to change the
organization from the top down then he would not even begin to become
involved with that organization.

Steve Delph

====================================================================






[Next of thread]             
[Next by author]            
[First of thread]            
[First by author]