Done. Quality confirmed that the restrained tick was a setup check only. The released hole-position reading remains a genuine failure against the applicable drawing.

That preserves the useful failure without inventing a measured change. Does the same bracket have a correction and reinspection record?

It has a final accepted hole-position result and a fabrication correction entry. The identity stays linked throughout. Correction duration is missing.

Then you know the route, not its labour allowance.

We followed a current selected group with the process specialist. Their first released checks were recorded before correction, and the observer separated hands-on work from waiting. The estimator is using that observed group, not the old tick, for the provisional allowance.

Keep the sample size and selection with that allowance.

Did the observed group include first-time passes as well as brackets needing correction? That matters when the estimate is spread across an ordinary batch.