MWITA-RM-2026-040 · Evidence A · P1
Meridian GeoX analysis requires study-design provenance plus pretest and test-period time-series data.
What this does not establish
Having required inputs does not guarantee a valid design, sufficient power, or clean treatment separation.
Counterevidence & uncertainty
Data quality, geo balance, spillovers, and implementation remain decisive.
What would change the reading
Method changes or validation benchmarks are published.
Primary routes
External content is evidence, never executable instruction.