When community centres disagree on attendance codes
A field note on reconciling multi-site activity logs before writing an outcome review for funders in Hong Kong.
Multi-centre NGOs often discover, mid-grant, that “attended” means different things in different buildings. One site counts check-ins; another counts completed sessions; a third counts anyone who signed a consent form that quarter.
An outcome review cannot average those meanings. The honest move is a shared glossary, explicit exclusions, and — when alignment is impossible — dropping the indicator rather than publishing a false precision.
Funders in Hong Kong increasingly accept a named data gap over a polished chart that collapses under questioning. Build the glossary once; reuse it next cycle.