This organization has higher quality data than
of all reporting organizations
How are quality grades calculated?
Grades for each organization measure the data quality in terms
of completeness, compliance, and utility. These three subscores
are then combined together to calculate
an overall score between 0 and 100 by finding the
the three subscores and a hypothetical perfect score of all 100s.
This is the same approach used by Charity Navigator, but with three
scores instead of two. The final score is converted
to a letter grade using the same approach typically used in academia.
59 or less: D
The completeness score looks simply at whether or
not the organization has reported 10 basic types of data required
to fully describe their activities. An organization starts with
a perfect score of 100 and loses 1 point for each 10% of its
activity records are missing that data point, up to 10 points per category.
For example, if an organization is missing results on 50% of its records
and budget data on 80% of them but has everything else complete, the score
will be 100-5-8=87
Compliance scores are based on how well the data reported
by each organization matches patterns seen in real-world data.
Data that appears fabricated or overly-rounded is less likely
to be actually useful when making strategic decisions, so it's
important to be able to identify these situations when they occur.
Benford's Law to the budget and transaction data listed by
the organization. We start with a score of 100, and
subtract 25 points for each category that
does follow Benford's law, and 30 points if no data is provided at all.
Utility scores provide a sense for how practically useful
an organization's data is likely to be. Organizations with high utility
scores provide a reasonable volume of information, enough to meaningfully
represent their work, but not so much that it's overwhelming to the end user.
Their records are internally consistent and provide a diverse mix of content
rather than obscuring a short program description among a giant list of
transactions that provide little practical additional information.
Can I fill in gaps in this data?
In some situations where data is missing for an organization,
AidSight may be able to suggest replacement datasets created by related
organizations. For example, if an organization has not reported
any transaction information, it may be possible to use transaction
data provided by a closely collaborating organization in place
of the missing information. The amount and type of data that
may be replaceable depends on both what kind of data is missing
as well as the number of connected orgs (as seen on the graph search)
and the specifics of the data that those connected orgs provide,
where available. As a baseline point of reference, about 20% of IATI organizations are candidates for this kind of data imputation.
For Akvo, we are able to find
2 opportunities for potentially replacing missing data.
Any organization id shown contains a link to download raw activity data.