Publications with many more citations in OpenAlex than in other databases

[Gabor Schubert]

Hi,

I was doing a citation analysis for Stockholm University publications and found a few articles which have much higher "cited by count" in OpenAlex than in other databases. The most extreme case was a 2019 article from Scientific Reports (https://openalex.org/works/w2983201012) with 1787 citations as of today in OpenAlex, but only 30-40 in other databases (Scopus, WoS, Dimensions, Crossref). If I look at the citing articles in OpenAlex they are mostly from unrelated topics. The article is a psychology article, and the citing articles are from other areas than psychology. Most of the citing articles are from 2020-2021. I checked some of the citing articles and they don't have this psychology article in their reference lists. So obviously this is caused by some kind of reference matching error in OpenAlex, although it is unclear what the exact origin of this error is.

I tried to find other similar cases in OpenAlex among the highly cited publications in general, and I found an Indonesian article from 2024 (https://openalex.org/works/w4402690901) with more than 2700 citations. The weird thing here is that most of the citing articles were published earlier than the cited article. According to Crossref this article hasn't received any citations yet (https://api.crossref.org/works/10.30640/dewantara.v3i2.2644)

Has anyone seen similar discrepancies in OpenAlex citation counts?

Gabor Schubert,
Stockholm University

[Nick Haupka]

Hi Gabor,

I found similar cases for paratext-like publications in OpenAlex:

For example, if you search for „Table of Contents“, you will find several publications with a high number of citations, although most of these publications are just ordinary tables of contents. 

Example: https://openalex.org/works/w4242011962

I also found some examples of publications that “cite” these tables of contents, but also have a publication year that is earlier than the actual cited publication. 

Example: https://openalex.org/works/w4239391401 (From 2016, but is cited from an article from 1998)

Best,

Nick

[Gabor Schubert]

Hi Nick,

Interesting. Although both of your examples have exactly the same citation counts in Crossref as in OpenAlex:

32 citations for w4242011962: https://api.crossref.org/works/10.1071/HC14001

10 citations for w4239391401: https://api.crossref.org/works/10.5771/2193-0147-2016-4-u1

in this latter case all 10 citing articles are from SSRN Electronic Journal. I checked one of them and it has really weird references in the original article: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=65028 and also in Crossref: https://api.crossref.org/works/10.2139/ssrn.65028 , but the DOI for 10.5771/2193-0147-2016-4-u1 is actually there.

So in these cases the error is not specific to OpenAlex, but probably stems from an error in the citing articles somehow.

Best,

Gabor

[Gabor Schubert]

I see now that both of your cases are actually originating from errors in the citing articles' original metadata. So these are not errors in OpenAlex. The "Table of contents" with DOI 10.1071/HC14001 is actually cited by some articles, at least this DOI is given erroneously in their references. See for example: https://bmcmededuc.biomedcentral.com/articles/10.1186/s12909-020-1963-6#Bib1 , where reference #68 has the erroneous DOI 10.1071/HC14001 (if you click on "Article" after the reference).

Gabor

On Wednesday, 26 February 2025 at 22:15:39 UTC+1 nick....@gmail.com wrote: