posted June 13, 2005 09:46 PM               

(Please feel free to correct)

My understanding is that Six Sigma is a quality philosophy rather than a measurment technique. The quality philosophy assumes that the process actual quality can not be truly measured by statistical techniques. So while 3 sigma techniques would seem to indicate the process limits where 99.7% of all product would fit into +/- 3 sigma, reality has taught otherwise.

Six sigma uses the observed statistical quality numbers, makes them worse than measured, and then compares that result to the specification. If the six sigma quality range turnes out to be wider than the specification, then "Six Sigma" process improvement techniques are unleased on the process in order to have the processes "six sigma" numbers fit into the specification.

As part of the philosophy everything can be measured and it's true variation can be estimated and improved upon with the "Six Sigma" program.

Personally I don't have an opionion whether this is good or bad or better than all the other philosophies I seen over the years. If you get perople to pay any attention to quality, that is often better than the usual.

OK, have at it.

------------------
Best Regards,

Tom Cunningham

Extrusion Technical Services

posted July 09, 2005 08:23 PM               

The +/- 6 Sigma allows a manufacturer to get rid of all the junk that overfills the +/- 3 Sigma bell curve.

Pick a sigma number, say 0.0018. Now, from the center line (zero), multiply this # three times for your plus and minus spread.

posted July 09, 2005 11:02 PM               

Sigma (1 standard deviation) defines the shape of the curve by mathematical formula. How many times you multiply it does not change the shape of the curve. The number of standard deviations from the mean can be used to estimate the portion of the population defined by the bell curve when truncated outside that number of standard deviations. In other words more of the population, when sampled, will fit under +/-6 sigma than +/-3 sigma.

As stated previously the 6 sigma philosophy assumes the sampled populations variation to be worse than normal calculations. While the actual variation of the population is fixed, 6 sigma atrificially makes sampled variations worse than calculated in order to make it more difficult for the sampled population fit within specification.

To my understanding, the effect is the opposite of what you are saying.

------------------
Best Regards,

Tom Cunningham

Extrusion Technical Services