Research
Measuring outcomes, not attendance
A seat filled for ten weeks is not the same as a skill learned. The M&E framework was rebuilt around the second thing, and it made the numbers less impressive to report.
It is much easier to report attendance than outcomes. Attendance is a headcount; outcomes require an assessment model that can distinguish a student who can build a working circuit from one who sat through ten sessions about circuits. We had the first kind of number for a long time before we had the second.
Building the second kind meant assessed outcomes at learner, class and school level, tied to specific curriculum modules rather than a general 'engagement' score. It also meant reporting figures that are less flattering than an attendance count, because assessed skill acquisition is always lower than seats filled.
The M&E framework we now use for public-sector partners is built the same way — baseline, midline and endline design, with the methodology published rather than summarised. A partner who wants to check our working can, and we'd rather field that question than avoid it.
The uncomfortable part of this shift is internal, not external: it means some programmes that looked successful by attendance looked considerably less so once measured by outcome, and we had to say that out loud in our own reporting before asking anyone else to trust the numbers that replaced it.
A seat filled for ten weeks is not the same as a skill learned. The M&E framework was rebuilt around the second thing, and it made the numbers less impressive to report.
Working on something similar?
If this post raises a question about your own deployment, that's a conversation worth having directly.