News

How we think about the problems clients bring us.

Written by the consultants. No confidential client work appears here.

LatestJuly 2026 · 6 min

What an evidence plan actually needs before you run the first test

Most testing programmes are designed backwards. Decide what claim you need to defend, then work out what data defends it.

Mateus RochaRead more

Reviewers form a view in the first half page. Three aims that depend on each other read very differently from three that do not.

What our 2025 study found, and what it means for how restorative dentistry is taught.

General models answer confidently and cite nothing. For clinical questions that is the wrong trade.

A strong record can still read weakly. The order and framing matter more than most applicants expect.

Specimen preparation, storage, and reporting choices that quietly invalidate an otherwise sound study.

Working software and defensible software are different standards. Only one survives review.

Tell us what you are working on.