How each track is built

Each track is anchored to the regulatory and professional body frameworks governing its own target population. No cross-sector borrowing.

• Review of the professional body AI guidance specific to the track, such as ABA Formal Opinion 512 for legal, FSB and IOSCO for financial services, NACD and OECD for board leadership, or C2PA for creative and media.
• Competency domains derived from those standards rather than from a generic template.
• Target population and exclusion criteria defined before any item is written.
• Literature review of existing measurement instruments in the field.
• The theoretical construct documented and agreed before drafting begins.

Subject matter experts draft 40 items per track, calibrated to observable professional behavior rather than to abstract knowledge.

• 25 behaviorally anchored rating-scale items distributed across the competency domains.
• 15 scenario-based judgment items concentrated in the highest-stakes areas.
• Three proficiency levels per item: Acquire as the safe minimum, Deepen as competent applied practice, Create at expert and governance level.
• Anchors describe what practitioners at each level actually do, not what they know in the abstract.
• Reviewed by the track subject matter lead before it reaches the expert panel.

An independent panel scores every item on relevance and clarity. Nothing enters a published instrument without passing this gate.

• A minimum of six subject matter expert judges per track, independent of the item authors, following Lynn (1986) and Polit and Beck (2006).
• Every item rated separately on two criteria: relevance to the construct, and clarity of wording.
• Item-level (I-CVI) and scale-level (S-CVI/Ave) indices computed on the four-point ordinal scale established by Lynn (1986).
• Acceptance threshold for the full instrument: S-CVI/Ave at or above 0.90.
• Items below threshold on either criterion are revised, not removed, and resubmitted for a second independent round.

The validated instrument is integrated into the platform and the track opens to organizations.

• Scoring rubric and level thresholds locked. No post-publishing item change without re-validation.
• The track becomes available to organizations holding an applicable access code.
• Response data begins accumulating from licensed professional respondents.
• Volume monitored in preparation for the psychometric stage.

As responses accumulate, the instrument is tested formally for factor structure and internal consistency.

• Exploratory Factor Analysis once 150 or more responses per track are available.
• Confirmatory Factor Analysis at 300 or more responses, to confirm the domain factor structure.
• Internal consistency measured against a Cronbach's alpha target of 0.80 or above per domain subscale.
• Item performance and score drift monitored between formal cycles.
• Periodic refinement as the professional and regulatory landscape moves.