Label their conditions and explain the applicant’s role. Distinguish a simulated result from actual deployment performance.
Preserve the boundary of the evidence
If the test excluded a major operating condition, say so. A factual account can still demonstrate useful knowledge; hiding limitations can make the proposed duty explanation unreliable. Identify who designed the test protocol and which variables were deliberately controlled, since a report that only lists outputs from a template generator says little about the applicant's own analytical role.
Where the applicant interpreted ambiguous or conflicting results and decided how to proceed, describe that interpretation directly rather than letting the raw numbers stand in for the judgment behind them. Distinguish a report the applicant personally authored from one the applicant merely reviewed or countersigned, because a countersignature alone does not establish the depth of knowledge the position requires. Across revisions of the same test, note what changed in the method and why, since a static snapshot can hide the iterative reasoning that actually demonstrates specialized knowledge.
Keep the description tied to what the applicant can explain in their own words rather than to the polish of the final formatted report. Outside reviewers or peer engineers who commented on the report can show whether their feedback changed the applicant's approach, since a documented exchange demonstrates the applicant actively defending or revising a technical position rather than simply filing a result.