Powering Potential
These ratings describe how dependable the extraction is from the current files. They do not judge whether the underlying program work happened. This is the quick read on which sources are ready for management use and which still need cleanup before they support longitudinal analysis and future collection.
The implementation timeline remains the strongest cross-school history in the archive. The important interpretation change is that “training” is usually logged as an activity attached to another implementation, not as a standalone project type.
Each bar shows the number of schools with deployment or upgrade activity in that year. Hover over a bar to see the underlying row counts, overlap with training, and the deployment labels driving that year.
This isolates training labels that do not also appear with a deployment or upgrade label in the same year. Hover over a bar to see the training rows, overlap, and the labels behind that year.
These are the non-training implementation labels that most often show up in the same school-year as training. This is the clearest evidence that training behaves like a companion activity.
Older outcome coverage is stronger than the recent files alone suggest. The exam archive is the biggest structured student-outcome layer, and the older graduate survey provides row-level alumni responses that can later be normalized into a reusable outcomes table.
This is the strongest legacy outcome source because it preserves school-by-year pass-rate history and Form V qualification rates across multiple clusters.
These are raw respondent counts from the older row-level survey workbook, not final organization-wide percentages.
Useful but narrow: one school cluster, three years, and a 2023 schema change that requires a comparable measure rather than a straight raw total.
These are the substantive questions the current archive can already support, even before a full cleanup is finished. The point here is not perfection, but what we can already learn if we organize the strongest historical sources well.
These are the source layers that are most worth turning into clean master tables, standard definitions, and repeatable collection instruments. This is the practical bridge from historical cleanup into future M&E.