Monitoring baseline selection record
StatMedAnalysis | 9 November 2026
Use aggregate or synthetic process observations.

1. Set a measurement unit and the time or acquisition grid. List planned exclusions before calculating limits.
2. Enter one row per candidate interval. Supply an operating-log reference for each suspected change.
3. Freeze the chosen interval and the decision for future rebaselining.

Process / measured unit: ____________________  Cadence: ____________________
Interval | Instrument/version | Conditions | Number of occasions | Known discontinuity | Include and why
________ | __________________ | __________ | ___________________ | ___________________ | _______________
________ | __________________ | __________ | ___________________ | ___________________ | _______________
Chosen interval / baseline version: __________  Owner and decision date: __________
Rebaseline trigger and approval record: _________________________________________
Fictional example: 30 weekly synthetic benchmark runs span a software update after week 18. Mark weeks 1–18 and 19–30 separately. Examine comparability and uncertainty within each segment before choosing a reference. Combining all 30 by default can hide a change in level.
Completion check: every excluded observation can be traced to the predeclared rule or a separately documented exception.