All historical data should have, if possible, observation-level attributes attached which indicate the set of processing steps applied, or at least the status of the observation, e.g. whether it is passed through directly or has been adjusted/harmonized by the item code.
This follows #51, since the SDMX community has already established code lists (with the ID "OBS_STATUS") for this purpose. See:
All historical data should have, if possible, observation-level attributes attached which indicate the set of processing steps applied, or at least the status of the observation, e.g. whether it is passed through directly or has been adjusted/harmonized by the
itemcode.This follows #51, since the SDMX community has already established code lists (with the ID "OBS_STATUS") for this purpose. See:
These include precise codes like:
E: Estimated valueK: Data included in another categoryM: Missing value; data cannot existWe should understand and try to apply these, only inventing new codes if necessary.