Notes
Overall results are typically reported as an average value. In addition to the
average value, we also want to report the repeatability of the measurements.
One way to report repeatability is by using Relative Average Deviation.
Relative Average Deviation
Relative Average Deviation (RAD) is expressed as a percentage and essentially
tells us how much repeated measures fluctuate. A RAD of 1% would mean we
expect each trial to differ by about 1% from each other. This is effectively the
uncertainty associated with our measurement. A smaller RAD means your
measurements have higher precision. Here we will use a four-number data set
(97, 99, 100, and 103) to demonstrate how to calculate RAD:
Step 1: Calculate the average of all your trials:
Step 2: Take the absolute value of difference between each trial and the
average:
Step 3: Calculate the average of the four deviations from the average found in
Step 2:
Step 4: Calculate the RAD by taking the average of deviations in Step 3, dividing
it by the average from Step 1 and multiplying it by 100 to get into a percentage.
20