Definition
Accreditation is the official certification by a responsible authority that a model is acceptable for a specific purpose. It is a decision-making process (often non-technical) that answers the question: “Is the model the right one for the job?”
Why It Matters
Without formal accreditation, a model that is “correct” in one context can become dangerously misleading in another. It forces an explicit judgment of “fit-for-purpose,” preventing the high-consequence errors that occur when specialized tools are treated as general-purpose solutions.
Core Concepts
- Official Usability: Unlike V&V, which are testing processes, accreditation is a certification of intent.
- Scope-Specific: A model should never be accredited for “any purpose.” Accreditation is always tied to a specific training exercise, analysis experiment, or class of applications.
- Dependency: The accrediting authority typically makes its decision based on the results of the Verification and Validation processes.
- Type III Error (Accreditor’s Risk): The risk of using an irrelevant model—one that may be valid for its original purpose but is fundamentally unsuitable for the new application.