Evidence · Sources · Method
Every claim, traced to its source
Most people-analytics content asserts. This one cites. Below is the full proof graph: each claim we make in public, linked to the graded evidence behind it — a peer-reviewed corpus, citation-grade findings, a source-graded registry, filed patents, and original research. When an AI assistant or a skeptical reader checks our work, it holds up.
The evidence base
Counts are computed from the repository, and the 20 evidence links below are verified on every build (npm run proof:check). No claim ships unsourced.
The proof graph
People decisions are the most consequential decisions organizations make — and the least rigorously measured.
People analytics is not a dashboard — it is behavioral science and statistical inference applied to workforce decisions.
The academic literature is 50+ years deep — and almost entirely out of reach for the practitioners who need it.
- Corpus 205+ profiled and searchable works
- Original research Original research bridging the literature to practice
AI is being deployed in high-stakes people decisions without the measurement science to evaluate whether it works or whom it harms.
- Insight card Citation-grade findings on AI risk/validity in people decisions
- Original research Original AI-human-interaction research
- Encyclopedia AI × people-analytics capability encyclopedia
PeopleAnalyst is the missing manual: peer-reviewed methods, citation-grade evidence, translated for practitioners at the pace of AI adoption.
- Insight card 40+ citation-grade insight cards
- Corpus The translated corpus
The evidence is indexed and searchable: 205+ works, 40+ citation-grade insight cards, 8 research arcs.
- Corpus 205+ works
- Insight card 40+ insight cards
- Research arc 8 research arcs
Good measurement has a method: construct validity, reliability, and effect-size interpretation are what separate evidence from assertion.
- Principia registry The source-graded, citation-verified organizational-science registry
- Insight card Effect-size / validity findings
Our methods are original, not aggregated: filed patents and pre-registered research stand behind the engines.
- Filed patent Protected-feedback provisional patent
- Filed patent Consensus-coder (reliability-scored extraction) provisional
- Filed patent Adjusted-benchmark-index provisional
- Original research Pre-registered research protocol
Why this is hard to copy
Anyone can publish opinions about people analytics. The defensible position is the one underneath: a corpus read at chapter fidelity, findings extracted to citation grade, a registry that grades its own sources and carries Bayesian priors, and methods original enough to file. Source superiority is not a tagline here — it is a graph you can walk, a check that runs on every build, and a bar we hold the whole portfolio to.
The method behind the grading lives at Principia; the findings at Insight cards; the corpus at the Library.