Standards Day: Outline Agenda

[bill_anderson]

We propose the following themes and issues for the Standards Day. Each issue will be covered by its own paper. If you wish to add a new issue comment here, but if you wish to propose content within an issue go to the topic for that paper.

Rules and Guidelines

• Revised rules for decimal and integer upgrades (Discussion) (Paper)
• Deprecation of old versions of the standard (Discussion) (Paper)
• Codelist management (Discussion) (Paper)
• Hierarchies (Discussion) (Paper)
• Double counting (Discussion) (Paper)
• Traceability (Discussion) (Paper)
• Miscellaneous rules and guidelines (Discussion) (Paper)
• Controlling Data Quality (Discussion) (Paper)
• Results

Improvements to the Standard and Publishing Statistics

• Core funding and funding from own sources (Discussion) (Paper)
• Improvements to the Organisation Standard (Discussion) (Paper)
• Improvements relation to both Activity and Organisation Standards (Discussion) (Paper)
• Improvements to the Activity Standard (Discussion) (Paper)
• Improvements to publishing statistics methodologies (Humanitarian Paper)

#IATI #TAG2017

[TimDavies]

We’re doing some work on linking contracting and IATI data.

I hope to have a paper on options for this, including in all likelihood one or two proposed codelist additions or updates, by mid-Feb.

Is it possible to add this as a topic for the agenda?

[bill_anderson]

Hi @TimDavies

Absolutely

Do you see this as something resulting in proposals that can be included in the next decimal upgrade (i.e. it is optional and backwardly compatible). If yes, then it belongs on this day.

If your scope is broader or longer term or the proposed changes need to wait for an integer upgrade I would suggest we fit this into the Day 2 and 3 Agenda

[TimDavies]

Hey Bill,

At this stage I think the proposal on schema/codelists we would make
would be backwards compatible changes that allow optional linkage between
aid and contracts data over the coming year, but I think there might also
be interest in the longer-term discussion of whether or not this is
something to recommend more strongly in future - which could make for a day
2 set of discussions.

[bill_anderson]

I agree that two separate approaches are good. Could the broader discussion go a step further and look at how open data standards should, in general, seek interoperability?

[Herman (Herman van Loon)]

Hi Bill,
In line with the IATI mission and vision, I suggest to add a separate Data quality track to the TAG 2017 agenda for day 4. The Data quality of IATI data determines the usability of the standard in practice and is therefore critically important for the survival of IATI as a means for information exchange. In the last couple a number of applications have been developed by several people. The TAG would provide a great opportunity to share in depth the experiences about working with IATI data to produce meaningful information and what problems are encountered due to data quality problems.
Sharing these experiences could be starting point for practical proposals for improving IATI data quality on the near future. Workshops could cover the following topics:
• Experiences with IATI data quality from the data users perspective (e.g. country cases, OIPA development, digital activity progress monitoring with IATI) an identifying the key issues with data quality (both technical and organizational)
• Action to be taken by publishers, IATI technical support and data users to improve data quality thereby making sure those IATI goals as described in the IATI mission and vision document are met.

The topics double counting and traceability could also be part of this track. I also added this as a comment to the draft agenda.

Regards
Herman

[bill_anderson]

Hi Herman

Day 4 is planned as a single track with an agenda that focuses on reaching decisions (consensus) on a range of specific issues relating to changes to the standard, rules and guidelines. Double counting and traceability are key items on this agenda.

Broader discussions on data quality should definitely be on the agenda on Day 1 (demand side) as well as Days 2 & 3 (supply side).

Does this make sense?

[Herman (Herman van Loon)]

Hi Bill,
I was hoping to have an in depth technical discussion on the most common data quality problems with fellow ‘techies’ and see if we can come up with suggestions to improve/enhance the automated data validation process (currently supported by the XSD ) and the feedback process to the publishers.

I was thinking of day 4 because that seems to be the day for the more IATI technical discussions. Days 1-3 seem to be targetted at more broader audience.

[bill_anderson]

Ok @Herman, I’ll find a way of fitting it in to Day 4. It would be good, in a dialectical spirit, for the discussion to be framed around some specific ideas. Are you up for doing this?

[VincentVW (Vincent van 't Westende)]

@Herman @bill_anderson Totally agree on the need for a tech discussion on improving validation. With all the cross-publisher use cases that are starting to be implemented, the data quality indeed starts to undermine usability more and more while it seems the ‘techies’ think alike on improvements.

@Herman I’d be happy to provide input or feedback on framing.

[Herman (Herman van Loon)]

@bill_anderson I will be happy to do some suggestions and @VincentVW and others yes please provide input and feedback.

[dalepotter]

@Herman @VincentVW I’ve been also thinking about changes to the publishing process with the aim of improving data quality.

Shall we set-up a separate thread to share thoughts?

[Herman (Herman van Loon)]

@dalepotter, @VincentVW Yes, fine with me.

[pelleaardema]

@dalepotter, @VincentVW, @Herman Count me in too. I’ve proposed a session to share experiences on the earlier days too. Happy to do that in a more tech setting working towards solutions.

[bill_anderson]

So, what I think we need is some specific ideas regarding, for instance:

• What should be validated?
• How should it be validated?
• What happens with invalid data?
• How should the Registry handle invalid data?
• How should the Dashboard handle invalid data?
• How should Publisher Statistics handle invalid data?
• Should good quality data be ‘kitemarked’?

NB If there is a need for discussions on tools, support, etc these belong on Day 2 and 3

[dalepotter]

@Herman @VincentVW @pelleaardema @bill_anderson - I’ve set up a new thread: Technical measures to improve/incentivise better data quality

[danmihaila]

I think we should also define what invalid data means. If it is data that is not accesible (server is down, etc.) or something else .

[bill_anderson]

Here is our first take on an agenda for Standards Day (Day 4). We are now working on a detailed breakdown which is built around decision points: concrete proposals to make changes to schema, codelists, rules and/or guidelines.

Please note that the whole of the day is taken up with these decisions. Any other discussions need to be dealt with elsewhere.

Comments most welcome

[bill_anderson]

A special request to all those planning to attend Standards Day. For us to optimise the workspace please could you …

1. Sign in to the Sched App (and do everyone a favour by adding your picture and bio to your profile).
2. If you plan to participate actively please tick yourself as an attendee for all the sessions you may want to contribute to.
3. If you plan to attend primarily as an observer please only tick the Standards Day: Introduction session. (This doesn’t stop you from contributing.)

Can I remind everyone again that the aim of Standards Day is to try to reach consensus on specific technical issues relating to the standard’s schema, codelists, rules and guidelines. It is assumed that all participants will have a fair working knowledge of the standard. The agenda is long and we won’t have time to explain basic issues to novices.

[bill_anderson]

The decision points for todays Standards meeting are available here