Closed amy-silcock closed 4 years ago
Hi @gtkChop,
To help us decide whether we would like to pursue this idea, please could you provide us with the following statistics:
1) How many publishers have updated their data in the past 12 months? 2) How many publishers have updated their data in the past 24 months?
Thanks, Josh
Total active Publisher: 1321 Total Publisher Published so far: 1047
Publisher Count whose package metadata modified in past 12 months (from 2018-09-11): 844 Publisher Count whose package metadata created in the past 12 months (from 2018-09-11): 361
Publisher Count whose package metadata modified in the past 24 months (from 2018-09-11): 1053 Publisher Count whose package metadata created in the past 24 months (from 2018-09-11): 639
Hi @gtkChop,
Thanks for providing those stats!
I assume that these lists would include deleted publishers and publishers who have private files etc.
In which case, please can you update the counts to include only publishers which currently have an active file (i.e. are in the existing main page list - https://iatiregistry.org/publisher) Thanks!
@gtkChop This has become a high priority so would be great if stats can be obtained before our next call. To be clear, the qualifying criteria are:
Count 1
Count 2
@JoshStanleyDevInit Looking at this.
Here be dragons
By "package metadata modified" I understand that could be any part of a publisher corpus that has changed (overnight). Conceivably, a publisher could change ''last-updated-datetime'', but no other data element or attribute, and be considered as modified - and therefore active.
The IATI dashboard has a clear methodology in terms of what is considered to be an update (using the 'transaction' element). This makes the update metric more difficult to manipulate
The wider point is that there is a continued dialogue and discussion around timeliness - which is based on the IATI dashboard. Should new figures and lists (based on a different methodology) be distributed and disseminated - there's a significant risk of organisations questioning the validity of one or another
For display purposes, I do not understand why the metrics in the dashboard do not suffice: http://publishingstats.iatistandard.org/timeliness.html
If the cut off was Annual, then 758 publishers could currently be considered as "active"
@JoshStanleyDevInit
1102: Number of active organization (all active publishers. Includes all active publishers, who never published a dataset but active.) 641: Number of organizations with at least one active dataset within a year (from today).
Production numbers would be:
1372: Number of active organization (all active publishers. Includes all active publishers, who never published a dataset but active.) 1065: Number of organizations with at least one active dataset within a year (from today).
Time period - 1 year from the current day. Active Publisher: Published/Updated the dataset within 1 year and it's active (either organization file or activity file).
@JoshStanleyDevInit
Added active publisher count according to the description in an email that is, all active publishers with at least one public activity file.
Split the count according to the time frame:
Counts reflect the same as production (taken latest data dump from production).
Please see the staging site and let me know your feedback. Thanks
@JoshStanleyDevInit @Ocre42
You can generate the report through UI. I have sent an email with link to staging.
Thanks @gtkChop. Having it available on the admin screen has been super helpful. I'll close this for now and we can reopen if we opt to add a count to the public site in the future.
As we reach 1000 publishers we would like to change the display to say how many 'active' publishers there are. These are organisations who have updated their data within the last year.
Potentially, the new figure could replace the old one. Or we could have the two running side by side?
@gtkChop how feasible this and how much time would this take to implement?