Capability Analysis - Unusual Statistical Distribution of my Proccess

J

Jparfer

Hi there!

I have a big problem with my process. I want to see the distribution that my proccess have, but its not belong to a known distribution.

How can i stimate capability with less than 100 data using not distribution supossion?

Thanks!
 
T

t.PoN

Re: Statistical Distribution of my proccess

Use histogram to figure out what is your distribution.
the ASQ has nice excel sheet to help you doing your job. i have attached them below, i hope they can help

To have the capability study file and the distribution, please click here
 

Attachments

  • data-point-histogram.xls
    72 KB · Views: 226
  • SPC.xls
    1 MB · Views: 469
Last edited by a moderator:

Bev D

Heretical Statistician
Leader
Super Moderator
Are you required by your Customer to report a Cpk or Ppk number? Or are you genuinely intereseted in your process capability?

what is your process? what feature, property or function are you measuring?

The best way to understand your process capability to plot the data in time series agianst the spec limits. A simple run chart if there is only value per part or a multi-vari if there are multiple values per part. This is the starting place and the most informative. once you understand the nature of yoru process you can proceed to histograms and control charts...

If you post your data, we can provide more specific help
 

Steve Prevette

Deming Disciple
Leader
Super Moderator
Also, you can still use SPC to determine if the process is stable and for the basis of capability calculations as SPC is non-parametric.
 
S

supreecha

Re: Statistical Distribution of my proccess

:confused:Comparision Sheet by Excel , Minitab & Crytal Ball .
 

Attachments

  • Cpk.xls
    1 MB · Views: 451
  • Ppk.xlsx
    1.9 MB · Views: 415
Last edited by a moderator:

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
Attached is distribution analysis that shows clearly that you do not have a normal distribution. That said, we know that the regular Cpk and Ppk calculations may not be applicable.

The real question is: should this characteristic be non-normal? Does it make sense? The only way to determine that is understanding the process. Since we do not have any idea what the process is, very little can be discerned about whether it is stable or in control at this point.
 

Attachments

  • cpk.docx
    77.2 KB · Views: 186
Top Bottom