How Do I Get the ZBench Value?
Six Sigma – iSixSigma › Forums › General Forums › Tools & Templates › How Do I Get the ZBench Value?
 This topic has 8 replies, 2 voices, and was last updated 8 years, 8 months ago by MBBinWI.

AuthorPosts

October 3, 2012 at 12:56 pm #54197
Joe RobertsGuest@jrobertsz66 Include @jrobertsz66 in your post and this person will
be notified via email.Hi, I am trying to compute the ZBench value based on my results and I am getting an * for the ZBench value in Minitab. The only reason that I need it is that is part of a school project which requires the ZBench value.
Basically, I am getting these results:
LSL = 0
USL = 10
Target = 5
Sample Mean = 30.2516
Sample N = 20
Std Dev (within) = 0.089
Z.Bench = *
Lower CL = *
Z.LSL = 339.45
Z.USL = 227.24
Cpk = 75.75
Lower CL = 99.83
Upper CL = 51.66I would appreciate it very much if someone can explain why I am getting a * for the ZBench value in the Minitab.
BTW, I am a programmer and not a statistics guy so all this terms above are new to me. I have no clue what Cpk or CL is but I believe that ZBench is basically the sigma level, meaning how many DPMO I should expect.
Anyway, any help that you can all provide will be appreciated very much.
Joe
0October 3, 2012 at 1:03 pm #194159
Joe RobertsGuest@jrobertsz66 Include @jrobertsz66 in your post and this person will
be notified via email.BTW,
Just in case, here is my sample data:
30.135
30.375
30.152
30.278
30.168
30.153
30.375
30.177
30.136
30.315
30.285
30.355
30.339
30.351
30.252
30.232
30.351
30.238
30.175
30.1890October 3, 2012 at 1:09 pm #194160
Joe RobertsGuest@jrobertsz66 Include @jrobertsz66 in your post and this person will
be notified via email.Also, is this correct?
ZBench = Min(((Mean – LSL)/Std.Dev)), ((USLMean)/Std.Dev)))
0October 3, 2012 at 3:29 pm #194161
Chris SeiderParticipant@cseider Include @cseider in your post and this person will
be notified via email.Nope. Hint: Take a closer look at your USL and LSL and where your data’s central tendency is.
0October 4, 2012 at 5:29 am #194167
Joe RobertsGuest@jrobertsz66 Include @jrobertsz66 in your post and this person will
be notified via email.Hi Chris,
Thanks for your reply. I think what you are saying is that because my range for LSL and USL is 0 – 10, and because my mean is normally 30, then that explains why ZBench is *. Is that correct?
If so, then I don’t understand why I also get a * for ZBench with these values:
4.768
4.766
4.736
4.836
4.820
4.789
4.758
4.788
4.851
4.789
4.758
4.801
4.773
4.721
4.805
4.789
4.745
4.805
4.774
4.758Those are very close to the middle of the range 0 – 10, yet I still get an * for the ZBench value.
Here are my results from minitab with the 4.x values:
LSL = 0
Target = 5
USL = 10
Sample Mean = 4.7815
Sample N = 20
StdDev(Within)=0.0359276
StdDev(Overall)=0.0326456For Potential within Capacity, I get the following figures:
ZBench=*
Lower CL=*
Z.LSL = 133.09
Z.USL = 145.25
Cpk = 44.36
Lower CL = 30.26
Upper CL = 58.47For Overall Capacity, I get the following figures:
ZBench = *
Lower CL = *
Z.LSL = 146.47
Z.USL = 159.85
Ppk = 48.42
Lower CL = 33.30
Upper CL = 64.35
Cpm = 7.36
Lower CL = 6.96Thanks in advance,
Joe
0October 4, 2012 at 5:54 am #194168
Joe RobertsGuest@jrobertsz66 Include @jrobertsz66 in your post and this person will
be notified via email.Or are you saying that because all the points are so close to each other then it is impractical to calculate the ZBench value because the chance of getting a single DPMO is very slim?
0October 4, 2012 at 6:26 am #194170
Joe RobertsGuest@jrobertsz66 Include @jrobertsz66 in your post and this person will
be notified via email.HI @CSeider,
Not sure if you are getting my replies or not. Not sure how I am supposed to make sure my replies get back to you.
If you have a moment, I posted some additional follow up questions, as I am not sure that I understand your hint completely.
Thanks,
Joe
0October 4, 2012 at 4:20 pm #194174
Chris SeiderParticipant@cseider Include @cseider in your post and this person will
be notified via email.Contact Minitab if it’s that important to understand the mechanics of their calculation specifics but the last set of data is so narrowly distributed and well within both specs you have no issues–both Zu and Zl are superb. If this is real data, find a new problem to work on.
0October 12, 2012 at 3:40 pm #194212
MBBinWIParticipant@MBBinWI Include @MBBinWI in your post and this person will
be notified via email.@jrobertsz66 – now that your test is over, perhaps others may learn from this thread. In your first data set, you are essentially asking for the ratio of infinity over a small range (in your case, 10). In the second, you are asking for essentially zero over a range that is proportionately very large (even though the same range of 10). The calculated numbers, therefore, are either extremely large or small, thus Minitab defaults to a null instead of giving a nonsensical number (although Joel might have more technical justification).
0 
AuthorPosts
You must be logged in to reply to this topic.