Do the gage capability studies on BOTH kinds of rocks. Reason: the variability is very likely to be proportional to the mean. Therefore if you combine both kinds of rocks in the same study, your variability estimates will be too high for the small rocks and too small for the big rocks.That said, what kind of ROCKS do you manufacture?
Thanks for your determination on getting me straight on this. I found it. I had tried different search phrases earlier without luck. Perhaps I did stumble across the correct section but didn’t realize it at the time. Thanks again.
The link still just gets you to the “front page” of the handbook. I searched under non normal capability but couldn’t find the appropriate discussion. But, thanks for providing the equation. That helps a lot, but I would have liked the actual link as well. When I get a chance, I’m going to compare this equation to the software that I…[Read more]
Thanks for the link, but I couldn’t find the alternative calculation you mentioned. Could you provide the complete link, rather than just the home page? I’d like to compare it to my old DOS-based software.
I don’t know where these other guys are coming from, but Cpk reflects short-term variability. By definition, Cpk uses RBar/d2 in the denominator, not standard deviation. RBar reflects within subgroup variability. By definition, RBar measures short-term variability. Standard deviation however (the one with n-1 in the denominator) refl…[Read more]
To get Ppk instead of Cpk, simply substitute the sample standard deviation (the one that uses n-1 in the denominator) for RBar/d2 (“Sigma-prime”). It’s the same for Pp, Ppl, and Ppu.
Ken, I bet you already know the answer to your question, but here is my (quick) answer anyway.
Short term: RBar/d2 (“sigma-prime”).
Long term: the sample standard deviation (that’s the one with n-1 in the denominator).
I hope no one has any complaints with this explanation. It’s not very detailed, but if you think about it, most processes…[Read more]
I tend to agree with Ken on just about everything except one point. He stated “D. Wheeler and others, including Montegomery say control charts are effective regardless of how the data is distributed. In essence, the control charting tool is robust or insensitive to the distribution of the data.” I’ll agree with this provided that the…[Read more]
Go to http://www.iqtsixsigma.com/.
They are in the middle of a Six Sigma class now. Send them an e-mail asking for the next available class schedule.
The instructor is Gary Steele and he is excellent. Of all the quality and statistics instructors I’ve had (almost 500 classroom hours not counting grad. school), no one rates above Gary.
Some folks in my company have developed what looks like a pretty good evaluation matrix for Six Sigma black belt competency. There are 14 categories that are assesed on a scale of 1 to 10 then weighted according to the importance. The target is to get at least a “7” in each category. It’s in an Excel ’97 file and I’d be glad to e-…[Read more]
Ken, thanks for your response. I don’t like 2-sigma limits either, but in this case (automatic retesting) I think we agree that they are OK, especially when talking about something as important as calibration.
Your logic of point number 3 (on the one day approach) makes sense. Your initial post however confused me when you stated the AVERAGE…[Read more]
I tend to agree with Ken with one exception. After your control chart is made and implemented, I’d add a 2-Sigma “warning limit” IN ADDITION TO the regular 3-sigma limit. If any point exceeds the 2-sigma limit, I would do a retest. If this retest also exceeded the 2-sigma limit, or even came very close to it, I would do the re-calibration. Rea…[Read more]
Stay away from the “low hanginf fruit”. You don’t need Six Sigma tools to solve those problems. On the other hand, don’t pick a problem with a low probability of success either.
Take a hint from Paul Stone (above) and take your controller to lunch. He/she should have some ideas how you can save in excess of $100,000 per year.
Doesn’t your s…[Read more]
I’d like to see the e-mail addresses return. Reason: I’ve e-mailed information (mostly attachments) to posters in the past and they were appreciative of the information. If e-mail addresses cannot return, then you ought to consider giving us the ability to attach documents. The easiest way to handle this is to bring back the e-mail addresses.
Whoops, you are right. There is a whole chapter in Grant & Leavenworth. It has been at least 17 years since I looked at the book, except when I checked the index for ‘Rational Suggrouping”.. Guess what? No mention of it in the index. How can he devote a whole chapter to it without mentioning it in the index? Beats me. Thanks for setting me st…[Read more]
99.997% reliability is equal to a Sigma Level of 5.51 (assuming the usual 1.5 Sigma Shift). Here is how you can calculate it using Excel:
1. Compute the proportion defective. In your case it is 1 minus 0.99997 or 0.00003.
2. Substitute in this formula using Microsoft Excel:
Doing so will yield a…[Read more]
Here’s my understanding of “green belt”. Six Sigma “Green Belts” study much of the same material as “Black Belts” but in not as much depth. While BB’s generally take 4 weeks of classes and are expected to work full time on a project that can deliver in excess of $100,000 per year to the bottom line, GB’s take 2 or less weeks of classes, and…[Read more]
- Load More