“Zlt = Zst – 1.5”
This is the “Six Sigma” answer…and probably the one that your course is looking for. It is an estimate based on “empirical evidence compiled by the early adopters of Six Sigma based on on 34 years of statistical data…*blah, blah, blah*.
That being said, there are many “doubters” (myself included) that question whether this actually holds any water in “real life”. The assumption is that your process will be worse in the long term, because of “shift and drift” effects in the process. The concept itself makes sense, it’s the 1.5 that people tend to have a problem with. In fact, many practicioners have thrown this out the window.
My take: If you want long term estimates, take long term data. In the mean time, don’t assume that your process will be “as good as it is at the moment” over the long term. I also tend to believe that this “shift” will vary depending on where you are on the “curve”. Keep in mind: a 1.5 Sigma shift at 3 Sigma is a whole lot different than a 1.5 shift at 6 Sigma.
For example: By this calculation, if your short term capability is in fact 6 Sigma (2 defects per billion opportunities), then your long term will be about 4.5 Sigma (about 3.4 defects per million opportunities). Ok, that seems reasonable…I can see picking up some defects here from shift and drift.
But if you short term capability is 3 Sigma (1350 DPMO), then by this calculation your long term capability will be 1.5 Sigma…a whopping 67,000 DPMO! That I have a little harder time with…
This whole discussion is one of the “trees” that a lot of people have missed the “forest” for…
So your best bet is to ignore everything I just said. Just be careful…make sure you know the difference, and more importantly how you, your organization, and the people you talk to are coming up with these numbers. Ultimately, IMHO, it doesn’t matter whether you use the shift or not, as long as you’re consistent.
(By the way, there are about a million threads on this topic on this forum…I hope I didn’t just start another one…)
.
.