Report the observed range, uncertainty, sample scope, and method limits without forcing a winner or claiming equivalence.
Report the result before interpreting it
When a comparison shows no obvious difference, present the observed results under the shared test conditions and state that the available evidence does not support a clear ranking. This is more accurate than forcing a winner because small observed differences may reflect sampling variation, measurement uncertainty, or limits in the method rather than a dependable performance difference.
The presentation should identify what was compared, the conditions held constant, the sample scope, and the range or distribution of observations. If measurement uncertainty is relevant, it should accompany the result or be explained in terms appropriate to the method. NIST Technical Note 1297 offers general guidance for evaluating and expressing measurement uncertainty, but it does not supply evidence for any particular product.
Do not turn inconclusive evidence into equivalence
Insufficient evidence of a difference is not proof that two products are equal. Equivalence requires a defined acceptable margin and a comparison designed to evaluate that question. Without those elements, the proper conclusion is narrower: no clear difference was demonstrated by this dataset and method.
Sampling boundaries matter as well. The NIST Engineering Statistics Handbook notes that acceptance sampling primarily supports a decision about whether a defined lot is likely acceptable; it is not intended to estimate the lot’s overall quality. Likewise, a limited comparison should not be generalized beyond the represented samples, versions, or lots.
Make limitations part of the result
A transparent comparison names limitations such as sample size, incomplete controls, restricted test range, or uncertainty large enough to obscure a difference. It should preserve results that do not match promotional expectations rather than hiding them. The closing statement can identify what additional evidence would improve discrimination, while leaving the current outcome intact: the test did not establish a winner, equality, or superiority.