Can rejections of weak separability be attributed to random measurement errors in the data?

Standard non-parametric weak separability tests do not account for measurement errors in the data. We propose a method to determine if rejections of weak separability obtained using such tests can be attributed to random measurement errors in the quantity data.