The MML Framework

Psychological science has developed robust standards for evaluating measurement instruments — reliability, validity, and generalizability. It lacks a systematic approach for evaluating the measurability of the constructs themselves. The Measurement Maturity Level (MML) framework fills this gap: it grades the current state of measurement readiness for each construct, from well-established to novel.

MML is framework-agnostic — it can be applied to any psychological taxonomy, not only HES. It evaluates not whether a construct is valid or important, but whether it can currently be measured with the available scientific tools. A lower MML level does not mean a construct is wrong. It means the measurement infrastructure is not yet mature.

Levels are ordinal and cumulative: reaching a higher level requires satisfying the criteria of all lower levels. Not every construct must or should reach MML-1; some may remain inherently difficult to measure.

MML uses a structured six-step decision tree. Each step is a necessary condition for reaching the next level. Failing any step assigns the construct to the corresponding level.

1. Does the construct appear in peer-reviewed scientific literature? If NO → MML-5. If YES → Step 2.
2. Does a dedicated, validated instrument exist? If NO → MML-4. If YES → Step 3.
3. Does the instrument meet minimum reliability standards (α ≥ .70 or equivalent)? If NO → MML-3. If YES → Step 4.
4. Has the instrument been used in ≥ 5 peer-reviewed publications? If NO → MML-3. If YES → Step 5.
5. Is there evidence of convergent validity with at least one related measure? If NO → MML-3. If YES → Step 6.
6. Does the instrument meet the criteria for high-quality measurement — α ≥ .80, independent validation across samples, discriminant validity, normative data, and use in ≥ 20 peer-reviewed publications? If YES → MML-1. If PARTIALLY → MML-2.

Applying MML to all 343 HES elements yields the following distribution:

60% of HES elements sit at MML-1 or MML-2, indicating reasonably mature measurement infrastructure for a majority of the framework. 22% are at MML-3 (emerging measurement, opportunities for further validation) and 18% at MML-4 (recognized constructs without dedicated instruments — targets for instrument development). The absence of MML-5 entries reflects that all 343 elements were deliberately anchored to existing research traditions, even where the specific formulation is novel.

The distribution varies systematically by layer. Core elements (stable trait-like capacities) reach higher maturity levels more often than Quantum states (transient mode-shifts), which are harder to measure by their nature. Full construct-by-construct classification is in Supplementary Table S1.

Three reasons MML is a useful addition to psychological science:

Transparency. Constructs vary widely in how well they can currently be measured. Existing psychometric standards evaluate instruments; MML evaluates constructs. Making the difference explicit lets researchers, practitioners, and readers interpret empirical findings with appropriate confidence.

Research prioritization. By identifying constructs with theoretical importance but low measurement maturity, MML highlights where instrument development and validation would yield the largest gains. Conversely, constructs with high maturity but limited theoretical integration may warrant conceptual refinement.

Honest comparison. Without explicit measurement-maturity grading, comparisons across constructs carry implicit bias — as if all were equally measurable. MML makes this bias visible and correctable.

The full MML article — including the literature review, detailed methodology, discussion of measurement modalities, and construct-by-construct classification of all 343 HES elements — is available as a preprint on PsyArXiv and as a downloadable PDF.

Level changes are recorded in the changelog as the evidence base moves.

An MML level is a statement about evidence, not about individuals. It should inform how much weight to place on a construct's current measurement — not be read as a fixed or final verdict.