iSixSigma

Attribute Data Capability Study

Six Sigma – iSixSigma Forums Old Forums General Attribute Data Capability Study

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • #45871

    Wayne Bartscher
    Member

    Could anyone help me with a way to use Attribute Data and determine the Capability of a process? The process is parts being supplied to our facility and the suppliers are using Go/NoGo gauging on the parts. Our Design Release Engineers have set a 1.33sigma capability requirement on parts checked with only Attribute Data from our suppliers.

    0
    #150748

    LBJ
    Participant

    Wayne 
     
    What you are trying to do is actually very easy.  Simply take the number of no go items and consider that a defect, then take the total items checked as your opportunities.  Take these numbers and simply plug them into the Sigma Calculator on the toolbar above.  If you need further assistance just ask. 

    0
    #150749

    The Force
    Member

    If you’re using either Minitab or JMP — you just utilize the attribute process capability which will give you values related to ppm, z score BUT no direct Cpk or Cp since these are meant for continuous data only.

    0
    #150763

    Ropp
    Participant

    What crap !
    I’ll bet you also believe in the tooth fairy ?

    0
    #150793

    Robert Butler
    Participant

      Chapter 9 of Measuring Process Capability by Bothe is Measuring Capability with Attribute Data.  It is far too long to summarize here but it does show you how to derive equivalent process capability estimates using np charts.  I don’t know Minitab but based on the post by The Force it sounds like Minitab may do exactly what Bothe describes in his book.  I’d recommend reading the chapter and then taking a look at what is available on Minitab.

    0
    #150804

    Wayne Bartscher
    Member

    Thank you all for your responses I now believe I have the necessary information. Thanks again…..

    0
    #150907

    El’Dupe
    Participant

    Hello Wayne;
    Sorry for the belated response but it was a long weekend!  I read with interest the conversation’s development and have a couple questions as I am confused with respect to the process vs. result thing.  I’m going to back up a bit here and I hope I’m not off-base here but here goes…
    You mentioned the design folks want a 1.33 CPk min. and the supplier utilized a go/nogo gaging solution to determine acceptability.  This tells me that the data you have really won’t get the design folks the information they seek and my thoughts are because the data (go/nogo) clouds the real data (a measurement, a specific result) that would tell you what the process actually delivers (specific data within the design engineer’s specification tolerance.)  I have an adversion to measurement via hard pin for example, to get real data that can determine process capability.  After capability is established via a repeatable process plan I could see (even with a small 30 piece sample with specific results), but to use bucket information to determine capability tells me that the process isn’t critical enough to require all the effort being extended.
    For arguements sake:   For example, a go pin really only needs to contact an undulating hole wall in 3 places when inserted and to indicate non-acceptance.  The same applies to a nogo keeping the pin from being accepted in an oversize condition.  The hole could be actually oversized.  We know the shape of the hole could be oval, tapered, or out of perpendicular and accept hard gaging readily.  But…it hides the process’s real condition/capability and therefore reports false with respect to process capability.  This is no problem if the hole is not a major or critical item on the spec., but if it is, you will never really know via hard gaging.  I guess what I’m trying to ask is; what does the engineer really want/need and firming that up, is the supplier using the correct measurement technique to get it?  Right now I see a disconnect.  Thanks for listening and good luck!  See Ya!

    0
    #150919

    W W
    Member

    Bothe talks as much BS as Montgomery.  Read Wheeler.

    0
    #150936

    Paradise
    Member

    The tooth fairy must be real,!!!
    That is the only and the best way to run the attribute capability analysis, with the Sigma Calculator. Cummins Engine Conpany, who just happens to have surpassed the $1 Billion mark in validated savings due to our Six Sigma Program, uses the Sigma Calculator for all of our attribute data analysis….

    0
Viewing 9 posts - 1 through 9 (of 9 total)

The forum ‘General’ is closed to new topics and replies.