JSIFS Methodology Group · 2026-06-22
Null and negative runs are usually the first casualty of storage pressure and the last thing anyone writes up. They are also the material that determines whether an anomalous run can be interpreted at all.
Without nulls, an artifact audit cannot establish a baseline, a weak-signal analysis cannot budget false positives, and a replication packet cannot tell an external team what failure looks like. The cost of retention is small and fixed; the cost of absence is paid at exactly the moment a result matters.
Retention needs the same metadata standard as positive runs: instrument identity, acquisition context, operator, and deviations. A null run stored without provenance provides reassurance rather than evidence.