Data Quality Index Consultation
DATA COMPLETENESS
Sub-sections 2.6.1 - 2.6.4 Identifiers and Traceability
Instructions for submitting your feedback
1. Read through the proposed methodology for this measure and / or download the attached PDF at the bottom of this page;
2. Share your feedback through the comment box below, consider the guiding questions in your comments and include the question number in your response;
Proposed Measures - 2.6.1 IDENTIFIERS AND TRACEABILITY
Please find below the proposed methodology for this measure.
DEFINITION Asses if activities contain a Reporting Organisation Identifier that matches the IATI Registry
OUTPUT |
|
METHODOLOGY |
Count the number of active activities which have a reporting_org_ref that matches their IATI Org ID on the IATI Registry. Divide by total number of active activities. |
Proposed Measures - 2.6.2 IDENTIFIERS AND TRACEABILITY
Please find below the proposed methodology for this measure.
DEFINITION Assess if participating organisation contain organisation identifiers.
OUTPUT |
|
METHODOLOGY |
----------------------------------------------------------------------------
|
Proposed Measures - 2.6.3 IDENTIFIERS AND TRACEABILITY
Please find below the proposed methodology for this measure.
DEFINITION Assess if participating organisations contain activity ids
OUTPUT |
|
METHODOLOGY |
For active activities:
|
Proposed Measures - 2.6.4 IDENTIFIERS AND TRACEABILITY
Please find below the proposed methodology for this measure.
DEFINITION Assess if transactions contain provider and receiver organisation identifiers
OUTPUT |
|
METHODOLOGY |
For active activities:
For active activities:
|
Guiding question - please refer to the index number when you respond via the comment box below!
- Which transaction types should be captured?
GO BACK TO MAIN DQI-PAGE
Webinar
For each discussion, the IATI Secretariat will organise a webinar to explain the proposed methodology, answer questions and further explain how to engage.
Michelle, agreeing with you in general, but could you clarify which specific Index element you are commenting on? I'm only seeing a reference to the IATI Registry for 2.6.1 (and 2.6.2, but that seems to be a mistake).
Yohanna Loucheur - Thanks for asking because now that I re-read the proposal I think I was reading it incorrectly. Of course a reporting org should have its correct identifier from the registry. For some reason I read it that participating org id's had to all come from the list of reporting org IDs in 2.6.2 and 2.6.4 and my point was simply that many participating organizations (excluding the reporting orgs which by definition are publishers) are not IATI publishers so they wouldn't have IDs that match and a publisher shouldn't have a lower percentage because of it.
What I would propose is that if a participating org is not a publisher then the structure of the identifier used should start with the 2-digit ISO country code in which the organization is headquartered. I don't know if the system can check for that so it may not be feasible.
Michelle Levesque yeah, I think the best that we can aim for is that a "valid org ID" would be one that aligns with org-ID.guide guidance/lists (not sure how to formulate, but that would be the general idea). In most cases, this would indeed mean an org ID that starts with a country code.
Yohanna Loucheur and Michelle Levesque on the discussion between valid vs well structure Org refs we're trying to capture this in 2.6.2, having two outputs one that checks if the org ref matches one on the IATI Registry. The other looks to see if any org ref has been included, to the second we could expand the scope of the check to see if it matches a prefix in org-id.
I can see the benefit of this but I'm also wary as in IATI data we see examples of organisations adding additional variables onto the end of their own reporting-orgs. Either way the check can be 'gamed', would be good to hear your thoughts if checking against the Registry + checking against Org-id is a reasonable solution.
Amy Silcock I see the pros and cons. Yes publishers could game the system by simply putting in the first part of the org id structure (e.g. US-EIN or GB-COH) and nothing else and that isn't a real org ID. However, comparing to Org-ID prefixes does provide a minimum of quality that is missing today which I could see as a benefit. As I've mentioned a few other times in the context of IATI issued publishing numbers, ensuring the first two characters are based on where the org is and not where the registration organization is I believe adds quality and context to the IDs even if they aren't complete with real registration numbers. Checking that the first two characters actually represent a valid ISO 2-digit code is a positive check in my opinion.
I would propose that IATI technical scrape (I think the techies call it that) the registry for anywhere org ids are listed. Either they, a working group, or a temp consultant, could analyze the data to see where we have multiple IDs for same org, figure out which if any are actually correct and provide feedback to the publishers to see about correcting where necessary. And that same scraped data could be provided so that publishers would have it as a reference to use rather than have every publisher have to construct org ids from scratch for non publishers. This is a project outside the specific scope of the data quality index but I do believe would aid in improving data quality on this key element.
Amy Silcock
Amy, I'm afraid I don't understand what you mean by "in IATI data we see examples of organisations adding additional variables onto the end of their own reporting-orgs. Either way the check can be 'gamed' Since 2.6.2 is not about reporting-org, I don't see how this would play in.
In any case, the fact that we still don't have the actual methodology being proposed for 2.6.2 is the main problem. It's very difficult to comment on something we cannot see, there's a great risk of error.
At this point I am tempted to err towards caution and suggest to drop 2.6.2 until a methodology can be properly consulted. What is absolutely clear is that the IATI Registry cannot be the only source of "valid organisation ID", given that many participating orgs are not publishers. For instance, we just agreed on a way to refer to partner country government bodies; these should be considered valid Org-IDs, but would not be on the IATI Registry.