# More multiple DOI madness

**URL:** <https://community.crossref.org/t/more-multiple-doi-madness/4090>\
**Category:** Metadata quality improvements\
**Tags:** metadata, conflicts, content-registration, metadata-retrieval, metadata-quality\
**Created:** [21 September 2023 07:44 UTC](https://community.crossref.org/t/more-multiple-doi-madness/4090 "2023-09-21T07:44:56Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![rdmpage](https://sea2.discourse-cdn.com/flex020/user_avatar/community.crossref.org/rdmpage/32/643_2.png) [@rdmpage](https://community.crossref.org/u/rdmpage)\
**Post date:** [21 September 2023 07:44 UTC](https://community.crossref.org/t/more-multiple-doi-madness/4090/1 "2023-09-21T07:44:56Z")

</div>

Yet more multiple DOI madness (see [Duplicate DOIs keep being minted](https://community.crossref.org/t/duplicate-dois-keep-being-minted/3554) for more examples).

The journal _Science_, arguably one of the world’s leading scientific journals, has multiple DOIs for the same article in its back catalogue. For example, here are three DOIs for the same article (“Some New Generic Names of Mammals”): All resolve two a URL on the [https://www.science.org](https://www.science.org) site, but only one to an actual article (the others return a 404):

- [https://doi.org/10.1126/science.17.439.873](https://doi.org/10.1126/science.17.439.873)
- [https://doi.org/10.1126/science.17.439.873-a](https://doi.org/10.1126/science.17.439.873-a)
- [https://doi.org/10.1126/science.17.439.873.a](https://doi.org/10.1126/science.17.439.873.a)

The journal displays only one DOI on the page for this article, but the metadata is in CrossRef, so anyone doing searches for this article gets multiple hits, only one of which is correct.

It looks like _Science_ kept changing their minds about what their DOIs should look like when there is more than one article starting on the same page, but didn’t clean up after themselves.

Perhaps _Science_ can be invited to clean this up?

---

<div class="post-metadata">

**Author:** ![ifarley](https://sea2.discourse-cdn.com/flex020/user_avatar/community.crossref.org/ifarley/32/44_2.png) [@ifarley](https://community.crossref.org/u/ifarley)\
**Post date:** [22 September 2023 17:11 UTC](https://community.crossref.org/t/more-multiple-doi-madness/4090/2 "2023-09-22T17:11:29Z")

</div>

Hi @rdmpage ,

Thanks for flagging this to us! I’ve reached out to the metadata quality contact at the American Association for the Advancement of Science (AAAS), the owner/steward of the DOIs in question, to investigate and clean up.

More as I have it,  
Isaac

---

<div class="post-metadata">

**Author:** ![rdmpage](https://sea2.discourse-cdn.com/flex020/user_avatar/community.crossref.org/rdmpage/32/643_2.png) [@rdmpage](https://community.crossref.org/u/rdmpage)\
**Post date:** [22 September 2023 17:41 UTC](https://community.crossref.org/t/more-multiple-doi-madness/4090/3 "2023-09-22T17:41:03Z")

</div>

Hi @ifarley,

Thanks! I did some cleaning on the data that I downloaded, based on the assumption that the most recent DOI within a set of DOIs for the same article was most likely the “true” DOI. Based on a looking at a few DOIs manually this seemed to work. I’ve focussed on the articles 1880-1924, for which about a third of the 60,000+ DOIs are “bad”.

I hoping this will be an easy fix for Science to make 🤞

Rod