How can I calculate control limits for a parameter that is consistently increasing?

[#1]

I'm trying to establish control limits in a parameter that is consistently increasing.  I can do it in Excel with a running average calculation, but I haven't yet found a way to do this in JMP.  If anyone has found a way to do this, that would be greatly appreciated!

-Mike

[#2]

That is an interesting question. Control limits are used to establish whether a process is stable. It seems that you know that your process is not stable.

So I guess that you are okay with the fact that the parameter is increasing and you want to know if there is a change in the rate of increase?

A simple solution could be to use the difference between rows. You can easily create a column for this: Right-click on the parameter column header > New Formula Column > Row > Difference.

You can also create a moving average column in the same way, although I don't understand how this helps.

You might also want to look at the Range chart for your parameter. And also the CUSUM chart for parameter.

I have attached a simple example table with formula columns and saved scripts for different charts.

I hope that is helpful,

Phil

[#3]

One other idea I'll add to @Phil_Kay 's thoughts is this...when you wrote '...consistently increasing.' is that truly the case such that the measured characteristic just heads off into positive infinity? Or does the measured characteristic have a cycle with some known or expected frequency? If this is the case perhaps one other thought for you is some sort of time series based control chart method. Maybe an exponentially weighted moving average approach?