OCE.13:4.4 - Compare consequences under their supported basis

Put the qualified values beside their common basis and important differences. State what improved, worsened, remained unchanged or cannot be compared, and who receives each consequence. Keep losses visible even when the intended contribution improves.

Test the explanations that could alter interpretation: demand and case mix, staffing, seasonal conditions, other concurrent changes, selection into the observed set, delayed effects and changes to the measurement procedure. Do not list every imaginable rival. Seek a distinguishing result only when its answer can change the claim or decision.

A before/after contrast usually leaves these alternatives open. Report that contrast as such. Do not subtract unrelated scales into a net-success score or silently let a local efficiency gain compensate for a protected condition. A lawful trade-off, if needed, is a decision for the appropriate owner, not an operation performed by the observer’s spreadsheet.

If the receiving claim needs “caused by”, identify the exact causal-use question and obtain the applicable C.28 support and domain evaluation result. Different questions may call for different designs. Neither an appealing mechanism story nor the existence of a comparison group automatically supplies identification, validity or transfer to another organization.

A stronger causal study is not a universal prerequisite for a bounded response. A current assignment conflict or qualified protection result may support a specific authorized repair while the overall effect remains uncertain. Conversely, a favorable descriptive trend cannot authorize an action whose required conditions do not hold.