C Chart Issues - Posted Data and Charts in Post #7

[J PARTHASARATHY]

The guideline given in the AIAG SPC Reference Manual for c chart is as follows:
"Since the control limits are based on a normal approximation, the sample used must be large enough so that the number of subgroups with c=0 is small."
Please ensure you meet this; otherwise there may be some misleading inferences.

[Dr Dirt]

I posted the data in post #7 if anyone could take a look at it I'd be very appreciative.

If any other info is required I'll be glad to provide.

[Steve Prevette]

Fundamental problem is that you are not plotting counts - you are plotting defects per units. These are not c charts, so the c chart control limts (based upon the square root of the average) do not work. For defects per unit, you would need to go with u chart control limits. That is, the average plus 3 times the square root of (the average divided by the number of units per datum point).

[Dr Dirt]

Thank you very much everyone!!!!

I'm new to the forum, but plan on doing lots of reading and learning.

[statdoug]

I have not closely examined the data, but it sounds like you are charting a stable mixture here (Sampling from groups with differing proportions, but combining them to form a conglomerate proportion).
As an example, imagine two processes, one is running 3% defective, the other is running 97% defective. assuming we take the same number of samples from both processes, our percent defective for the combined data will be 50%.The control limits on .5 are at their widest. This is not an exact parrallel for your situation in that you have a lot more than two processes in your mixture. Still, I think it is the root of your problem. When charting a stable mixture like this, an Individuals moving range (IMR) chart is an effective solution.