Teacher-reviewed simulations are promising prototypes
Google Research describes a process for generating instructional simulations around teacher-approved goals. Its new public sample library focuses on middle- and high-school STEM.
Why it matters for leaders
A small curriculum-team trial could test scientific accuracy, accessibility and alignment. Require students to predict, manipulate and explain a result, then check transfer without the simulation.
The evidence
The technical report is dated September 16; the research announcement is September 17. Twelve U.S. teachers supplied 36 requests. Internal raters selected among five candidates per request; three requests failed the quality threshold. A separate expert-rating exercise used 40 requests.
Who was studied
Teacher evaluation of STEM simulations; no reported student efficacy trial.
What the evidence cannot tell us
The evidence concerns teacher usability and expert ratings, not student learning gains. Vendor authorship, candidate selection and limited teacher participation constrain generalization. The report identifies classroom efficacy as future work.
Read the original sources
Summary and leadership analysis from the Week ending 2026-09-18 brief. The edition’s selection notes explain source access and evidence limitations.