Ray's question about Six Sigma is pertinent as many statisticians quite
understandably think it means something completely different.  It does not
imply that some value is guaranteed to be within +/- 6 or 3 sigma of some
mean although there are statements that could be interpreted as such.  The
famous 3.4 parts per million claim that did a lot to damage 6S in the eyes
of many statisticians in it's early days derives from single sided 4.5
sigma where there is an unexplained and unjustified shift of 1.5 sigma in
the mean.  It is an example of business spin.

Six Sigma is a trademarked concept that is mainly about good procedures
but also incorporates some good statistical principles.  Rather than
working only at the 'foreman' level, it is marketed at the chief
executives who then believe that they can drive out uncertainty.  Such
beliefs of course are rather futile from a statistician's point of view
but with the belief come money and, when the statistical procedures show
that they work, rewards.  However each 6S programme is different to the
next and tailored to the target process or business.

I think we should not dismiss 6S because of it's tacky birth but should
attempt to improve it.  In fact it is my view that 6S is a step in the
right direction towards the use of advanced statistical modelling
procedures in the whole of business and industry.  You can start with
Shewart and move through all the industrial work by Yates (the first DoE
example was in 1936) into SPC, Taguchi, Six Sigma.  Why not jump to the
inevitable asymptotic solution?

Here in the UK we have been very backwards compared to the far east and
the US.  Technometrics is 45 years old now and shows that in the US, the
ASA and ASQ have been fully engaged.  The RSS has had an industrial and
business sections (QIF and BIS) but these have always been seen as the
poor relations to the model development and theoretical sides.  In fact
both sides need each other and these days with terabytes of data being
stored but not analysed, there should be rich pickings indeed for
statisticians.

So I welcome the RSS's involvement in a BS6S standard.

'Nuff sed agen!

John

John Logsdon                               "Try to make things as simple
Quantex Research Ltd, Manchester UK         as possible but not simpler"
[log in to unmask]              [log in to unmask]
+44(0)161 445 4951/G:+44(0)7717758675       www.quantex-research.com


On Mon, 18 Apr 2005, Ray Thomas wrote:

>
> >It really is time for this statistics industry to wake up and insist on
> >involvement.  The recent proposal for a BS Six Sigma standard is a step in
> >the right direction and the RSS will I hope be fully engaged there. (Don't
> >be afraid of Six Sigma - it is only a name and doesn't mean what it says
> >on the tin but does contain some good processes).  We need to look at the
> >way in which sematech pulled the US chip industry back - every chip plant
> >in the US must have a statistician and many have a department.  Intel
> >alone employs over 100 statisticians.
>
> About twenty years ago American colleagues told me that the US space
> programme used Japanese chips because they were more reliable.   Evidence
> published at that time confirmed substantially greater reliabily of Japanese
> chips.  Has Intel now caught up with Japan?
>
> The Japanese organised the world's first international quality control
> conference in about 1969.   The underlying message of the conference seemed
> to be 'we have transformed ourselves from being makers of shoddy goods into
> the manufacturers achieving higher levels of quality of conformance than
> have previously been achieved in human history - and this conference will
> tell you  how we did it'.    I don't think that Western observers believed
> that message.
>
> There was a lot about statistics in the conference but not so much about
> statisticians.   The message seemed to be that it is important to have the
> relevant knowledge at the *foreman* level where quality control is
> exercised.   Does Sigma Six limit itself to statisticians and statistical
> departments?
>
> The RSS makes noises about education the public about statistics - something
> that it does not seem to be very good at.   But has the training of shop
> floor workers in quality control methods ever been on its agenda?
>
> Ray Thomas
> 35 Passmore, Tinkers Bridge, Milton Keynes MK6 3DY
> Email: [log in to unmask]
> Tel/Fax 01908 679081
>

******************************************************
Please note that if you press the 'Reply' button your
message will go only to the sender of this message.
If you want to reply to the whole list, use your mailer's
'Reply-to-All' button to send your message automatically
to [log in to unmask]
*******************************************************