Date: Sun, 17 Aug 1997 10:53:36 -0500 (CDT) From: skat@flask.com To: Warner Losh <imp@rover.village.org> Cc: "Brian N. Handy" <handy@sag.space.lockheed.com>, Greg Lehey <grog@lemis.com>, "Julian H. Stacey" <jhs@FreeBSD.ORG>, chat@FreeBSD.ORG, platforms@FreeBSD.ORG Subject: Re: speed test Message-ID: <Pine.BSF.3.91.970817103223.29748A-100000@ns1.flask.com> In-Reply-To: <E0x06aN-0000i5-00@rover.village.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 17 Aug 1997, Warner Losh wrote:
> I've also seen it used, btw, to descibe extremely huge male sex
> organs. As to why it doesn't also describe things on the trailing
> edge, I know not the answer to that question.
>
> Warner
>
Description of three-sigma is context sensitive. Normally, researcher
will specify if he means one-tail or two-tail statistics.
In the case of Motorola example, they are talking about two-tail, since
they want the middle part of bell curve. The two extreams, too small or too
large, are rejected parts and there is no reason to differentiate.
In your male sex organ example, they are talking about one-tail.
for one-tail
1-sigma = 84.1% or 15.9%
2-sigma = 97.7% or 2.3%
3-sigma = 99.8% or 0.2%
for the two-tail you double the second number, which becomes the
percentatge of extreams, e.g. 1-sigma = 31.8%. The center portion of bell
curve ("Norm") becomes 68.2%
1-sigma = 68.2% or 31.8%
2-sigma = 98.9% or 1.1%
3-sigma = 99.9% or 0.1%
I believe Motrola was shooting for 6-sigma, which means less than
1.973176w-07% rejection rate. One reject in a lot of 5,000,000 parts.
Shin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.91.970817103223.29748A-100000>