growth-traits-lab studies which characteristics of a person actually drive development, and what makes them grow. Some are general — clarity of thinking, initiative, regularity, focus — the kind that matter no matter what you're working on. Others are specific to a craft: what moves someone forward in cooking isn't what moves someone forward in game theory, and the general set doesn't cover that by itself. Both get the same treatment: naming candidate characteristics, working out how to observe them, and testing which methods genuinely shift them. The lab runs independently of the team building the development program itself, so its findings can't be read as marketing for that program. Right now it's early: one method, one open pre-registration, and no results yet — just an invitation to join a small research cohort and help find out.
This is a research lab, not an audit or a product review — it isn't testing anyone's platform or grading anyone's tool. It's asking a research question about human development and being honest about how little of it is settled yet.
methodology/discipline.md— the three rules every finding in this lab follows: pre-registration before analysis, explicit evidence-rung labeling, no jumping from "this correlates" to "therefore do this."methodology/six-point-checklist.md— what a candidate characteristic has to clear before it's tracked at all.domain-traits-process.md— how characteristics specific to one craft (not the general set) get chosen or designed, with a worked example across three very different crafts.cohort/protocol.md— how someone joins the research cohort, and what happens once they do.cohort/publication-gate.md— the privacy policy that governs what this lab will and won't publish about real people, enforced automatically in CI.lab/— the analysis scripts, kept generic (no production credentials, no hardcoded schema).
No calibration results are published yet. The first iteration found that the data pipeline itself wasn't ready — some data sources didn't have enough history, some were misconfigured. So far, this lab's most concrete output has been diagnosing its own instrument, which is a legitimate first finding, just not the kind anyone can act on yet.
If you want to be one of the people this research is about — sharing enough tracked activity
data to help calibrate what these characteristics actually mean — open an issue. Enrollment
follows the protocol in cohort/protocol.md: consent first, a dry-run of your specific data
sources second, then a four-week warm-up before you're part of the confirmatory analysis.
MIT — see LICENSE.