An equivalence test asks whether the data support an effect small enough to be unimportant in practice. Researchers define an acceptable range of differences in advance, such as a narrow band around zero for a productivity measure, and test whether the true effect is plausibly contained within it.
This differs from an ordinary statistical test that asks whether an effect is distinguishable from zero. A nonsignificant result may occur because the study was too small or the estimates were noisy; it does not establish that two work arrangements perform equally well.
A common approach is two one-sided tests, or TOST. The researcher tests the lower and upper equivalence boundaries and may conclude practical equivalence when the statistical evidence excludes effects outside the chosen range. The result depends heavily on whether those boundaries were justified before looking at the data.
When a study says that hybrid work did not harm performance, check whether it performed an equivalence test or merely failed to detect a difference. Look for the margin, confidence interval, sample size and outcome definition before interpreting the finding as evidence of comparable results.