Transparency
Reports, findings, and the things that did not work
This page exists before there is anything to put on it. That is deliberate: the publication commitment is made now, while it is still costly to keep.
What will be published here
- Every pilot finding, with its status label, sample size, method, and the observation that would falsify it.
- Negative and null results, at the same prominence and in the same format as favourable ones.
- Budget actuals against the published pilot budget, including overruns.
- Every change to the method, with the evidence or failure that caused it.
- Regulatory status updates as Tennessee review proceeds.
Current status
Pilot cohort
Not yet running. No student data exists, therefore no outcome findings exist.
Reports published
None. The first report will appear after the first cohort completes a full term.
Regulatory
In formation. Not a licensed school; operations subject to Tennessee review.
Standing claims and their falsifiers
Hypothesis — not yet tested
Pairing melodic encoding with Socratic recall may improve durable memory retrieval relative to conventional review. Falsified if matched-cohort delayed recall shows no reliable difference.
Pilot finding — provisional, n is small
Students given a question rather than an explanation produce longer and more self-corrected reasoning in written work. Unblinded and small-sample. Falsified if blind scoring of matched writing samples shows no difference.
The reasoning behind these claims sits in the method. The money behind them sits in the pilot.
Be told when a report is published
Including the reports that go against us. One email per publication, nothing else.
We store your email address and nothing else you did not type here.