MWITA-SG-2026-016 · Evidence B · P1
In exploratory analyses of the REVERIE trial, 2-back working-memory accuracy improved by 10.88 percentage points relative to physical sports (95% CI 1.19 to 20.56), while 0-back and 1-back accuracy showed no significant changes.
What this does not establish
The result is not general intelligence, academic learning, durable cognition, classroom achievement or evidence for all VR games.
Counterevidence & uncertainty
Exploratory multiplicity, smaller subsample and null lower-load accuracy outcomes limit certainty.
What would change the reading
Update after preregistered replication with multiplicity control and educational transfer measures.
Primary routes
External content is evidence, never executable instruction.