Where it lives. Every record is held in the Labora Rounds corpus (Supabase), tagged missing-dead-persons-research — 72 records with full provenance, drawn from the two source reports and deepened by hand this week.

No ranking of sources. Facts are labelled by kind — official record, study, reporting, advocacy, family — never scored against each other. The official record gets no premium; here it has earned the least. The people closest to the dead are treated as the witnesses they are.

We list everyone. Nobody is filtered out at the door for their race, or because the sourcing is thin, or because a viral detail turned out wrong. Deciding in advance who counts is the gatekeeping this project exists to refuse. Leads enter as leads, labelled honestly, so they are not lost.

Belief is the default. The official ruling and the family's account are stored together and shown together, with equal weight. Where a family, given real access, reached its own answer — the Harsch and Lewis families did — that answer is honored too.

What "read at the source" means. According to PubMed, the peer-reviewed ones trace to the Lancet study 10.1016/S0140-6736(21)01609-3, Adams et al. 10.1111/1556-4029.70102, and the CDC-data pieces in NEJM and Annals.

Whose it is. The data lives under Labora Rounds. Any series Dr. Connor publishes is hers, under her own name — not an institutional masthead. This is the evidence beneath the work, not the work itself.