Instrument development
and peer review
ALMA is a practitioner diagnostic tool. This page explains how the instrument was developed, where it currently stands in the peer review process, and what that means for how you should interpret and use its outputs.
What "practitioner tool" means
ALMA has been developed using established principles from organisational psychology, AI governance research, and EU regulatory requirements. It is designed to generate structured, scored evidence of human oversight capacity — evidence that is meaningful and actionable for practitioners navigating the EU AI Act and ISO/IEC 42001:2023.
However, the instrument has not yet completed formal peer review in an academic journal. Until that process is complete, ALMA should be understood as a practitioner tool, not as a psychometrically validated academic framework. Scores and findings should be interpreted in that context — as structured practitioner evidence, not as clinically validated measurements.
Development methodology
The ALMA instrument was developed through a structured, multi-stage process grounded in both regulatory requirements and practitioner insight.
Literature review and regulatory mapping
The five dimensions — Psychological Safety, Critical Engagement, Conscious Ownership, Growth Orientation, and Adaptive Flexibility — were derived from a structured review of AI governance literature, human factors research, and the specific human oversight obligations in the EU AI Act (Articles 4, 9, 14, and 26) and ISO/IEC 42001:2023.
Item development and expert review
The 50-item instrument was developed iteratively. Items were reviewed by practitioners with expertise in AI governance, organisational psychology, and regulatory compliance. Items were refined to ensure clarity, relevance, and coverage of each dimension.
Pilot testing with practitioner cohorts
The instrument was piloted with practitioner cohorts across financial services, healthcare, and public sector organisations. Pilot feedback informed item refinement, scoring calibration, and the maturity level thresholds (Developing, Emerging, Established, Optimised).
Scoring model and risk pattern development
The dimensional scoring model and five risk patterns (Silent Automation, Diffused Passivity, Frozen Governance, Lonely Vigilance, Confident Blindness) were developed from pilot data and expert review. They represent practitioner-observed patterns of governance failure, not statistically derived clinical categories.
Peer review status
The following table summarises the current status of each stage in the formal validation and peer review process.
| Stage | Status |
|---|---|
| Literature review and regulatory mapping | Complete |
| Item development and expert review | Complete |
| Pilot testing with practitioner cohorts | Complete |
| Scoring model and risk pattern calibration | Complete |
| Formal psychometric analysis (factor analysis, reliability testing) | In progress |
| Academic journal submission | In progress |
| Peer review completion and publication | Planned |
This page will be updated as the peer review process progresses. Last updated: March 2026.
What changes after peer review
Peer review completion will strengthen the evidence base for ALMA's commercial claims and enable more precise language about the instrument's properties.
Before peer review (current)
- ALMA described as a practitioner diagnostic tool
- Scores described as structured, scored evidence
- Dimensions described as structured governance dimensions
- Risk patterns described as practitioner-observed patterns
- No claims of psychometric validation
After peer review completion
- ALMA may be described as a validated practitioner instrument
- Scores may reference psychometric reliability statistics
- Dimensions may reference factor-analytic evidence
- Risk patterns may reference empirical prevalence data
- Published findings may be cited in commercial materials
Stay informed
If you are a researcher, academic, or practitioner interested in the ALMA instrument's development, peer review process, or potential collaboration, please get in touch. We will notify existing clients when peer review is complete and findings are published.
[email protected]