Hello!
Similar to Ivo Bleylevens, who asked this question here (3 Sept 2024) and did not get a response, I'm puzzled by the way OpenAlex treats institution paper counts.
As we're doing a comparison of OpenAlex institution profiles to WoS\Scopus profiles for an academic paper, the question of which counts are 'official' in OpenAlex is important for us.
The works_count property of exact same institution object
https://api.openalex.org/institutions/I114027177 gives different count: 258428
The web interface at
https://openalex.org/I114027177 also displays 258400 (due to rounding).
As Ivo also observed earlier, this exact count serves as a hyperlink using works?filter=authorships.institutions.lineage:i114027177 so when a user clicks on 258400, they see 259900 papers (259938 in API)
So, what is the correct way to count an institution's papers in OpenAlex? 257687, 258428 or 259938?
I notice that the web interface uses authorships.institutions.lineage for all breakdowns by institutions.
Does this discrepancy mean that (some?) child institution paper counts are not included in filter=
institutions.id and\or institution's works_count? If yes, why?
For I114027177 there are 4 children with the largest having >9000 papers.
https://api.openalex.org/works?filter=authorships.institutions.lineage:i114027177,institutions.id:!I114027177 gives 2252, so numbers add up to the authorships.institutions.lineage totals. The 2252 papers are mostly by a child
https://openalex.org/I4390039329 with proper lineage in authorships in a handful of works I have checked manually.
Switching to web interface by deleting 'api.' prefix from the url above gives zero, because filter=
institutions.id seems to be disabled.
Would greatly appreciate any clarification!
Best regards,
Ivan