Principia — overview

Principia is a comprehensive, source-graded survey of organizational measurement — predominantly the people side of organizations. The output is two-shaped: a book manuscript and a queryable database. They are produced from the same underlying registry; the book is one rendering, the API is another, and anything else (a portfolio research surface, a printed reference, an SDK for builders) is a third.

What it codifies

For each construct family Principia covers, the survey produces:

1. The construct itself — definition, theoretical lineage, measurement-model assumptions (reflective vs formative; unidimensional vs multidimensional; latent vs observed)
2. Instrument inventory — every instrument that has measured the construct, with item-level schemas (item text, response options, scale anchors, reverse-coding, scoring rules)
3. Reliability and validity evidence — per instrument, cited from the validation literature
4. Effect-size table — meta-analytic and primary-study links between this construct and its predictors and outcomes, with study-quality grades
5. Cross-cultural notes — where the construct travels, where it does not, what gets lost in translation
6. Operational notes for builders — how to use this construct in production research, common implementation pitfalls

What it is not

• Not original empirical research. Original work happens in sibling apps (Vela, the People Analytics Platform, Fourth & Two) and feeds back into Principia where appropriate.
• Not a textbook substitute. Textbooks teach; Principia indexes. The targeted reader has read at least one psychometrics text already.
• Not a meta-analysis service. Where strong meta-analyses exist, Principia cites them; where they do not, Principia surfaces the gap and may file an analytical sub-study (see Preregistrations).
• Not exhaustive on day one. The construct-family roadmap (see PROGRAM.md) is sequenced; coverage compounds; the book ships when the spine is dense enough to stand on its own.

Why this matters

Load-bearing measurement is unevenly distributed across organizations and disciplines. The same construct is measured five different ways across five different studies. Effect-size tables live in the chapters of one textbook and the appendices of another. High-quality instruments get reinvented in low-quality form because the original is paywalled or buried. Principia exists to give builders, researchers, and operators a single place to look — graded, sourced, queryable — and to give the People Analytics Platform a canonical measurement vocabulary it can subscribe to.

The portable claim: the principal-issues thesis says every domain has a load-bearing measurement set, and most domains are stuck because they have not named it. Principia names the measurement set for organizations. The methodology — source grading, statistical-metadata extraction into a shared schema, novelty verification before publication, queryable indexing — generalizes. The same shape would work for clinical psychology, educational measurement, marketing research, or any field where rigorous measurement is unevenly distributed across the working community.

How it relates to the rest of the portfolio

Principia is one of three consumers of the eventual @measurement/core shared package — the canonical measurement vocabulary that meta-factory has already drafted in packages/core/ and that the people-analytics-toolbox/hub will consume for its metric library work. The hub-and-spoke discipline:

• @measurement/core — canonical vocabulary (Construct, Instrument, Item, Measure, Model, EffectSize, Publication, StudyQualityGrade)
• Principia — the canonical registry (the rows; the surveyed construct families; the graded effect-size tables; the book draft as it accumulates)
• People Analytics Platform / toolbox — consumes the registry via @measurement/core-typed reads to power the metric library, AnyComp's compensation-fairness measurement, Calculus's metric definitions, and Reincarnation's adaptive-measurement work
• Meta-factory — continues to do book and paper extraction, populating Principia where it surfaces instrument-level evidence
• Vela — uses the registry to ground its adaptive-measurement and instrument-validation research

A sibling app that discovers a new instrument in the wild does not silently expand the shared schema; it files a Principia-side assignment to register the instrument. This is the same hub-and-spoke discipline the People Analytics Platform applies to anonymization, metric calculation, and segmentation. Consolidate once, consume many times.

What's next

This is the v1 scaffold (2026-04-30). The construct-family surveys begin once @measurement/core is extracted from meta-factory and the queryable database schema lands. Schema extraction is tracked as ASN-1013 in vela's meta-factory consolidation drive. First survey on the queue: engagement (the densest accumulated literature; serves as the methodology proof-of-method).

For the construct-family roadmap, see PROGRAM.md. For methodology — source selection, grading, extraction, schema discipline — see methodology.md. For current state, see PIPELINE_STATUS.md.