which is the right control chart
Six Sigma – iSixSigma › Forums › Old Forums › General › which is the right control chart
 This topic has 7 replies, 8 voices, and was last updated 17 years, 7 months ago by MW Richardson.

AuthorPosts

February 17, 2005 at 3:07 pm #38460
guys Im just starting out with this project of implementing SPC on the shop floor at my company. I am in charge of design, set up and implementation. I want to make sure we are using the control chart. Most of the time we will have about 50 sets of measurement data with 35 subgroups each set. I was wondering if XbarR is a good choice or XbarS. Can someone help me out here?
Thanks in advance0February 17, 2005 at 3:19 pm #115020
HarkanwalParticipant@Harkanwal Include @Harkanwal in your post and this person will
be notified via email.Xbar and R charts are typically used to track the process level and process variation for samples of size 8 or less, while Xbar & S Charts are used for larger samples.
If you will be working with 50 sets of measurement data Xbar & S Chart is a better choice.
Hope this helps0February 17, 2005 at 3:58 pm #115024Almost.
Xbar R charts are used if your subgroup size is 10 or fewer. Larger than 10 you can use the Xbar S.
If your subgroup size is 35 (let’s say 5) and you have 50 measurements, you would have 10 subgroups charted on an Xbar R chart.0February 17, 2005 at 5:09 pm #115032Straight from the mouth of MiniTab:
“Xbar and R charts are typically used to track the process level and process variation for samples of size 5 or less, while Xbar and S charts are used for larger samples.”
MiniTab (or other stat software) will show you how the values are calculated. Take the above with a grain of salt – think about HOW the limits are calculated and how sensitive you really want your charts to be. Make sure you understand the difference between S and R and how the control limits are calculated.0February 17, 2005 at 6:34 pm #115038All of these are just rules of thumb. Go look for yourself the acccuracy of Rbar/d2 vs. sbar.
You will find that the error when using sbar for samples of 2 is very big compared to rbar. Other than that the error is a wash. Period, end of discussion. Shewhart and his boys knew what they were doing when they came up with the values for d2 (and they got values for up to n=25 at least).
Conclusion, don’t use s when n=2. Otherwise it does not matter.0February 28, 2005 at 6:53 am #115483
Sezgin EzgiMember@SezginEzgi Include @SezginEzgi in your post and this person will
be notified via email.I would also like to mention CUSUM charts which i believe very effective.
0February 28, 2005 at 5:00 pm #115501KISS. You can use either one. Just know what your looking at.
0February 28, 2005 at 5:24 pm #115503
MW RichardsonParticipant@MWRichardson Include @MWRichardson in your post and this person will
be notified via email.You are asking very basic questions. There is nothing wrong with that. If that is the case, and you haven’t done so already, you should consider taking a basic course in SPC. That being said, you are definitely on the right track. I have seen many projects fail to deliver the results they should have because of a fundamental lack of understanding of the process’ central tendency and variation over time.
MW Richardson,
Sr. Q.A. Eng.0 
AuthorPosts
The forum ‘General’ is closed to new topics and replies.