There is a simple way to determine whether the results obtained from performance measures are accurate, and this is to ensure that the same performance results are reproducible and repeatable using a different but similar dataset.

Let's consider a practical example:

Consider a speed test for websites at a web design company. The same website speed-testing tool should always be used to ensure that all websites are tested against the same benchmark, which will, therefore, make the results from different websites comparable.

If the same tool is consistently used, the test can be repeated as often as required and the stakeholders using the measures can be comfortable that the measure remains constant and accurate.