Article View: bit.listserv.quality
Article #151Re: Capability & Stability Question /G'house
From: L. H. Garlinghouse
Date: Mon, 06 Jul 1998 00:00
Date: Mon, 06 Jul 1998 00:00
50 lines
3532 bytes
3532 bytes
> Date: Mon, 6 Jul 1998 01:31:03 -0400 ThomCycle@aol.com writes: > > I've been going round and round with some colleagues about capability index > > (Cpk) and their inherent usefulness/unusefulness during early stages of > > introduction of spc. The thought are Cpk does nothing to give us unique > > information about the process without the control chart and the histogram. > > Can we abuse or overuse Cpk? We have many other measures which give us > > (worker, management, customers) information about the ability of the process, > > do we need to add this tool in the early stages of our process management > > implementation? Do we need it internally if our customers aren't interested > > in it? I would like to here the opinion of others. ------------------ Generally speaking I say, "Yes," Cpk is a useful tool, even at the early stages of an SPC initiative. It can also be overused. I have found it useful to attempt to impress on people that a number is just a number, and if you do not have the accompanying variables chart in control over an extended period of time, you do NOT have a capability index. If the chart is in control, you may have an index of: Short Term Process potential; Long Term Process Potential; Short Term Capability or Long Term Capability - depending on the period covered by the supporting charts and some other things. Be careful if management starts demanding good Cpk's without understanding getting processes in control is a full fledged undertaking in its own right. If the chart is not in control, then at best you have simply a snapshot in time of the behavior of your process during the time you collected the supporting data.. All of that said, Cpk, regardless how flaky can be useful. It will tell you if re-centering your process will help. It can be translated to a parts per hundred defective rate (or a ppm if you are fortunate) and management seems to be able to understand that. It may tell you that you are closer to ppm than you thought, or that you have some real problems. I have seen lots of cases of adequate Cp and Cpk during studies, and high reject rates when going back to normal. This indicated to me a lack of discipline and other things which sometimes can be very important. I don't know how much overlap there is with what Cpk can tell y'all and what your existing reports can tell you - so in your case capability indices may be redundant. Hope this is helpful. << All opinions, statements, &c are my own>> L.H. Garlinghouse, C.Q.E. Pocahontas AR U.S.A. (870) 892-4586 ext 7659 garlinghou@waterlooindustries.com
Technical Details
Message-ID:
<70FF9A548B@poca.waterlooindustries.com>Path:
arc1-ca.pugleaf.net!unknown!not-for-mailReferences:
<393AB056238@waterlooindustries.com>Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 1159be,d1fc664c49f04ef6,start X-Google-Attributes: gid1159be,public From: "L. H. Garlinghouse" <garlinghou@WATERLOOINDUSTRIES.COM> Subject: Re: Capability & Stability Question /G'house Date: 1998/07/06 Message-ID: <70FF9A548B@poca.waterlooindustries.com> X-Deja-AN: 369137579 Sender: TQM in Manufacturing and Service Industries Discussion List <QUALITY%PUCC.BITNET@vtbit.cc.vt.edu> Approved: NETNEWS@AUVM.AMERICAN.EDU * References: <393AB056238@waterlooindustries.com> Priority: normal Comments: ******************************************************** Approved-By: "Bill Casti (System Admin)" <help@QUALITY.ORG> X-Received: from SpoolDir by WI-POCA (Mercury 1.30); 6 Jul 98 13:56:01 -0600 Organization: Waterloo Industries, Inc. Newsgroups: bit.listserv.quality