OCE.13:4.2 - Make the consequence questions observable

Look across operating results, contribution crossings, coordination, human conditions, customers and relevant surrounding Systems. These are places to discover a missed consequence, not six compulsory metrics. Select only the questions whose answers can change the receiving use.

For each selected comparison, make the following recoverable in ordinary language or an existing measurement account:

• the exact subject and eligible episodes, including which people, recipients, failures or diverted cases are excluded;
• the arrangement, exposure and other conditions under which each observation applies;
• the observation period, relevant delay and baseline or other comparator;
• the characteristic, unit or category and the measurement or collection Method;
• missing observations, uncertainty and the basis on which the readings can be compared;
• the interpretation the receiving decision needs and the claim the evidence cannot yet support.

For a late evidence-return proportion, for example, identify the eligible requested crossings, the agreed receiving deadline, the observed return time and the rule for cases still open at the end of the window. “Late” must mean the same thing on both sides of the proposed contrast. If the rule changes, recover a common basis from the underlying observations where possible or state that the values are not comparable.

Use C.16 for the exact measurement question, including the applicable model, calibration, scale and uncertainty. Reuse a qualified measurement procedure instead of inventing a new one to suit the desired conclusion. A survey response about burden, an observed missed deadline and an inferred workload mechanism remain unlike results.

Preserve subgroup and reach differences that could change the decision. An average among completed cases may say nothing about those diverted, abandoned or never admitted. Do not enlarge the observed population merely because the decision concerns the whole organization.