Working Papers
A paper every six weeks, not every year.
The Institute runs its research as a rolling pipeline of six-week sprints, and publishes on a cadence it controls. One-a-year is a pace set by an older world. This is set by ours.
01The Cadence
One paper every six weeks, ramping to one a month.
A single study going faster does not get you there. A rolling pipeline does. Three cohorts, each running back-to-back six-week sprints staggered by about two weeks, means something completes roughly every fortnight. Layer mechanism, methods and durability papers on top of that flow, and monthly becomes reachable.
One honest distinction holds the claim up. Artificial intelligence does not compress the biology: a six-week intervention takes six real weeks. What it collapses is the tail, the analysis, the reproducible code, the drafting and the figures, so the lag from a locked dataset to a finished paper drops from months to days. Preprints remove journal latency entirely. We commit to papers shipped, meaning pre-registered, preprinted and submitted. Acceptance runs on other people’s clocks.
- Months 1–3One sprint at a time while the machine is built. First paper around week six to eight.
- Months 3–9The pipeline fills and cadence tightens toward every three to four weeks.
- Month 9 onMonthly steady state, as templates mature and the long arm’s durability papers land.
02What We Publish
Several kinds of paper, from one body of work.
- Outcome papers
- One per sprint per cohort. The proximal result, reported on the sprint’s end date.
- Mechanism papers
- Whether early movement in the MABERT capacities predicts the outcome. The causal fingerprint.
- Methods papers
- Chiefly the MABERT capacity instrument being built and validated, and the no-control staggered design written up as a method.
- Rapid reports
- Multiple-baseline and single-case designs are publishable with small samples and replication, which is what fast cadence needs.
- Protocols and pre-registrations
- The pre-registration is itself an output, lodged before any data, and a Registered Report where the design is reviewed up front.
- Durability papers
- From the long arm, on its own clock, reporting whether change holds at three and six months.
03How the Cadence Stays Honest
Speed without slicing or fishing.
A fast cadence with no control group invites exactly one suspicion. It is designed out from the start.
- Each sprint is pre-registered with a single named primary outcome before data.
- One paper is one distinct outcome, cohort, mechanism or method. Never one result cut six ways.
- Every pre-registered outcome is reported, including nulls. A null is a paper.
- Open data and open code ship with each paper.
04The Series
The first outputs are in preparation.
The series opens with sprint pre-registrations, then the papers that report each sprint. The moment one is lodged, it appears here in full, with its status stated plainly, whether pre-registration, preprint, working paper or published.
We will not list studies that do not exist. The cadence above is a commitment, not a record. The record begins the day the first sprint does.
The evidence base is built in the open, on a cadence you can watch. Researchers and clinicians are invited to help build it.