Data Set Specifications vs Distributions

Site Feedback
1

We’ve been getting asked questions about when to use Data Set Specifications vs Distributions in Aristotle. There are conceptual reasons to use one vs the other. However, the main considerations about whether to use Data Set Specifications or Distributions are related to functionality in Aristotle.

I thought it would be useful to note the functionality we get if we use one of them that we miss if we use the other. The main things we notice are:

  1. Data Elements include a list of inclusions in Distributions in the Related tab whereas there isn’t a list of inclusions in Data Set Specifications. These linkages are covered in the Graph for Data Elements. One of our teams is currently creating metadata and is keen for this functionality to be available.
  2. Indicators allow explicit linkages from numerators, denominators and disaggregations to Data Elements and Data Set Specifications but not to Distributions. It would be very helpful if indicators could be explicitly linked to Distributions and Paths as it would enable related metadata (e.g. the indicators a data element is included in). It is possible to do linkages in the descriptive text however this means the related metadata isn’t possible.
  3. In general Distributions handle lineage better than Data Set Specifications.
1 Like
2

Hi Andrew,

These are really good points. Thanks for raising them.

I have added all of these to our roadmap to work on. I will keep you updated once we add or release the mentioned feature enhancements and fixes.

Cheers,
Shikha

1 Like