If using a codelist, the data must exactly match

(Matt Geddes) #1

The new [SDG guidance] :slight_smile: (https://docs.google.com/document/d/11vsqHc-mdAYUCqrrGCaCTFxu48oD3VvBVSlY9Q3ideo/edit) states that: “IATI recommends that if a publisher’s internal indicators do not directly match any SDG Indicator, they should choose one SDG Indicator that their internal result’s indicator best relates or contributes to.”

I think this breaks how IATI works and am very surprised that this is allowed as this means that data users cannot rely on the data being supplied by a publisher, exactly matching the codelist entry that it is marked as.

This means that if a user were to aggregate all the baselines for a particular SDG indicator, they would not know that in some cases, the baseline data provided, is actually for a different indicator, not the one they are expecting, and therefore their conclusions from using this data would be false.

IATI needs to find a way to use existing tools e.g. the <legacy-data> element, or a new element, to allow publishers to indicate whether the data they are supplying is from a similar, but not exactly the same indicator/sector etc to that described by the codelist which they have used to mark it. This could easily done by an attribute i.e. match=not_exact or similar.