ME.16:4.5 - Explain a changed result before selecting an adaptation

Use this continuation when an unexpected result, changed setting or proposed adaptation leaves materially different explanations that would change the next application. Existing observations and a supported local correction may already settle the decision; a new study is not required merely because some causal uncertainty remains.

1. Recover how the proposed contribution is expected to work. Connect the needed result to the function each relevant contribution serves, the concrete actions or form that supply it, and the conditions under which the proposed mechanism could operate. Read the target and the introduction strategy separately. For example, a rehearsal may be intended to develop recognition of a stop condition, whereas an expert checking the completed result may supply that recognition directly. Use C.28.CM and the relevant subject practice to construct or challenge this explanation. Keep assumptions and unknown functions visible. Preserving a function does not by itself preserve Method identity.

2. Derive observations that could distinguish the live explanations. Compare what should be seen if the relevant contribution was not delivered, was delivered but did not activate the proposed mechanism, operated differently in this context, or was supplemented by another contribution. Several explanations may hold together. Link observations of actual actions, the proposed intermediate change and the outside result for the same cases and relevant times. Include failed entry and cases where the result differs. A procedure-completion count alone cannot decide among these explanations. Assess fidelity only against an interpretable baseline and only when that comparison can change the decision; unknown essential functions or an intentionally flexible strategy may support a narrower account of what was done instead.

3. Select an attainable comparison. First ask whether existing evidence distinguishes enough for the receiving action. If a further inquiry is worthwhile, pass the competing explanations, expected observations, case differences, timing, available access and whole burden to RMP.2. Its selected design and limits, operationalized through RMP.3, become inputs to the ME.11 trial or other inquiry. Check separately whether the proposed application can be performed and whether the evaluation can obtain an interpretable answer. Specify which changes the design permits: holding a configuration stable and studying an adapting configuration answer different questions. If no useful comparison is attainable, return the supported course or unresolved alternatives with the limitation that matters now.

4. Combine observations through the explanation. Use the selected methods to recover what was delivered, how participants used or responded to it, the pertinent context and the resulting benefit, harm or burden. Where broad records show a pattern, selected observations or interviews can investigate the difference; where a particular account suggests a mechanism, look for its expected consequences in the relevant records. Repeated observation can reveal a changed contribution or context. Keep the same cases, periods and measurement meanings recoverable when combining results. Separate an observation, a participant’s explanation and the investigator’s inference; use C.28 for any resulting causal reliance.

5. Make each adaptation interpretable. Compare the actual change with its prior target or introduction-strategy baseline. Recover the affected action or component, proposed function, reason, decision-maker and authorization, conditions, decision time and actual implementation time when their difference matters. Keep coupled changes together when their contributions cannot be separated. Determine what supports a claim that the function was retained, changed or remains unknown. An adaptation description supplies neither that functional conclusion nor an effect estimate. Return changed reusable operations or conditions to ME.15 even if the intended function stays the same.

6. Return what the evidence changes. ME.13 receives the conditions and configuration actually examined; ME.14 receives comparable consequences, support demand and displaced work; ME.15 receives the semantic change with separately qualified function and evidence; ME.16 receives the basis to retain, refine, branch, replace or stop the target, its introduction, or their support. A favourable result despite poor delivery may support investigating another contribution; it neither confirms the proposed mechanism nor makes all useful work a failure. A new supported explanation or useful alternative may warrant development without contradicting the earlier bounded account. Revisit uses whose relied-on premises change through their governing patterns. Use ME.24 when the change affects a source-coverage or reconstruction claim.

The result is an explanation sufficient for the named next action, with its evidence and unresolved alternatives. It can be a bounded continuation, a changed introduction strategy, a proposed target variant, a useful refinement of the explanation, or a stop. Additional detail is selected for the decision it can change.