I'm currently working on a project where I'm trying to figure out if there exists a significant difference between measurement instruments for RF power measurements. I will measure multiple Bluetooth modules TX-power with 4 different instruments.
Now, let's say I measure one module on all 4 instruments. The spread of the TX-power measurements is 0.15dBm. The measurement uncertainties of the 4 instruments are 0.1dB, 0.2dB, 0.3dB, and 0.5dB.
Is there any way to combine the uncertainties of all the instruments? In other words, have one uncertainty value that represents all instruments. For example, if the combined uncertainty would be 0.2dB, the spread of 0.15dBm would be within instrument uncertainty.
I think what you're trying to accomplish is to test the hypothesis whever all of the measurements have the same value (up to noise).
To compute one number that quantitatively answers such question you could use likelyhood ratio test.
You could for example assume that each device gives the same value with the prescribed uncertainty and compare it to alternative hypothesis that one of the devices gives values from a distribution with different mean.
You can read how to accomplish this here: https://en.m.wikipedia.org/wiki/Likelihood-ratio_test