Assessing IATI Data Quality

When publishing to IATI, our members aspire to submit good quality and timely data that gives an accurate representation of the organisation's spending and income, and allows the users of the data to gain an understanding of key funding relationships, future strategy and a picture of the supply chain.

This page shows how we conduct a review of a member organisation’s data to ensure they are on the right track. It contains some questions that we ask when looking at the data, key links and a step-by-step guide that you can follow if you want to understand the process by which published data is assessed.

For an explanation of some of the complicated acronyms, please refer to the Key Terms for IATI page.

Questions

Does the NGO have both activity files, and an organisation file on the IATI Registry? Both are required.

How many activities does the NGO publish? This gives a picture of the size and regional/sector focus of the organisation.

Which IATI Schema does the data meet? Is it up to date? IATI v1.03 is the earliest version that is still valid.

Does the XML validate against the IATI Standard validator (link below)? This validator checks if the data meets the rules of the IATI Schema, for example, are the dates in the correct format?

How often does the NGO publish? Monthly updating is the best standard, while quarterly is also a good frequency to be held.

Does the data give a timely reflection of the NGO’s transactions? The closer the dates of past transactions to the present, the better.

Is the data forward looking? Does it allow data users to see how the NGO is going to spend in the future?

How does the organisation’s data look on websites with visualisations?

Does the data meet the DFID minimum requirements?

Does the organisation publish beyond the DFID minimum requirements?

Step-by-step guide

  1. Visit the IATI registry and identify your organisation as a publisher
  2. Open your publisher link, and identify if there are activity files and an organisation file. A single activity file can contain multiple activities and numerous transactions. An organisation file is like a profile, and is a required element of IATI.
  3. Underneath the dataset file name link, which will be in light blue, there will be a download link. Clicking on this should open the XML file into a separate tab. Copy this URL and go to the validator (link above). Paste the URL of the activity XML into the bar that is revealed once your click ‘Fetch file from web’. Does it validate? What schema is it using? If it fails validation, what errors are highlighted?
  4. Go to the dashboard and select your organisation. In the table of contents there is a section for data quality. Any files failing validation will be highlighted here.
  5. In the top banner of the dashboard page there is a section for publishing statistics. Clicking on this will reveal two more options – timeliness and forward looking. Opening these links will show you a table where you can scroll down and find your organisation, and you will be shown how often any part of your organisation’s data is substantively updated. Clicking on the time lag option will show you how up to date this data is. If you select the forward looking link, you will be given a table that counts the number of activities for this and the next two years that contain budgets.
  6. There are a number of websites that visualise IATI data. While they make this data more accessible for the pubic, it also means you see your data from the user-end. Search for your organisation and any projects you are involved in at DevTracker, UN Habitat and d-portal. Do they offer an accurate portrayal of the data published? What errors can you pick out here that were not made obvious by the steps above?
  7. Download the Excel spreadsheet from gov.uk that outlines the DFID minimum requirements. This is a more intensive process, but is the most accurate way of getting a complete picture of your data quality. Open up the XML of your data in a browser that keeps the category format, for example, Firefox or Google Chrome. Go through the data, looking to see if all the required fields are present and if the content meets the standard outlined in tab 3 of the spreadsheet.