I am having trouble figuring out how to express uncertainty for a measuring machine with somewhat known error (I figure if I take the worst case of the error range we see, I am covered).
It seems that the uncertainty calculation expects you to have removed known error (also called systematic error, also called bias), but in our current system, there is no good way to do that. So I would like to include the error as part of the uncertainty statement. I am getting a lot of resistance on doing it different than other calibration certificates that are out there where they state a single value such at" k=2 0.754%T".
In my case I want to state "k=2 0.754%T plus 0.271%T potential error".
So far I haven't seen the error scale smoothly to express it as a formula as is often seen on linear dimensions.
I also don't want to just add it in as the Systematic error is not properly part of K=2.
Ideas? All comments appreciated.
It seems that the uncertainty calculation expects you to have removed known error (also called systematic error, also called bias), but in our current system, there is no good way to do that. So I would like to include the error as part of the uncertainty statement. I am getting a lot of resistance on doing it different than other calibration certificates that are out there where they state a single value such at" k=2 0.754%T".
In my case I want to state "k=2 0.754%T plus 0.271%T potential error".
So far I haven't seen the error scale smoothly to express it as a formula as is often seen on linear dimensions.
I also don't want to just add it in as the Systematic error is not properly part of K=2.
Ideas? All comments appreciated.
