How can I evaluate uncertainty of measure starting from recovery data obtained analyzing spiked samples during method validation?

#1
I'm a researcher working in the field of analytical method validation. I'm studying how I can evaluate the uncertainty of measure starting from recovery data obtained analyzing spiked sample. I found in literature an application in which it's evaluated the reproducibility calculating the RMS between the pool of the recovery data and the bias assuming as reference value the 100% of Recovery%.
Anyone can suggests me another way to evaluated the incertainty of measure or if the approach described above can be applied?