Attributes SPC study - Attributive control (Go gage)

[tornilleriadeba]

Hi all!

I need to perform a SPC study for an attributive control (Go gage). The customer is requesting to use 300 parts.

I have Minitab 19, but I don't find how to perform such analysis. Is it a capability analysis with binomial distribution? I tried this option, but if all measurements are OK (no defective part), it gives an error.

Thanks for your help.

Best regards!

 

[Johnnymo62]

I haven't used Minitab for this, but normally you need to put failing samples in with good samples. Ford used to be 50 pieces with 3 out of spec parts included.

 

[Miner]

I haven't used Minitab for this, but normally you need to put failing samples in with good samples. Ford used to be 50 pieces with 3 out of spec parts included.

I think you are confusing the OP's question with an attribute MSA study.

@tornilleriadeba You are using the correct analysis, and Minitab cannot perform the calculations without some defective units. What you can do is run a One-Sample Proportion test using summarized data. Do not check the box for 'Perform hypothesis test' and click OK. Minitab will display the mean value for p, which will be zero, but will also provide the 95% upper confidence limit for p.

 

[Johnnymo62]

Oops. Sorry.

 

[Steve Prevette]

You could also work backwards from the Poisson Distribution, but should get pretty close to the same answer if np is reasonably large.

 

[Miner]

You could also work backwards from the Poisson Distribution, but should get pretty close to the same answer if np is reasonably large.

With n=300 the results should be virtually identical.