Retractions on Crossref

Retraction Watch retractions now in the Crossref API

Martyn Rittman — Wed, 29 Jan 2025 00:00:00 +0000

Retractions and corrections from Retraction Watch are now available in Crossref’s REST API. Back in September 2023, we announced the acquisition of the Retraction Watch database with an ongoing shared service. Since then, they have sent us regular updates, which are publicly available as a csv file. Our aim has always been to better integrate these retractions with our existing metadata, and today we’ve met that goal.

This is the first time we have supplemented our metadata with a third-party data source. Until now, our APIs have included metadata provided by Crossref members along with outputs from our internal enrichment workflows, such as matches found for bibliographic reference matching and funders. Third party metadata has been gathered in Event Data, but this has been stored and delivered separately.

Knowing when work has been retracted is critical for assessing the integrity of research, and this enhancement of the data will be a great benefit to the community.

Where does the data come from?

Retraction Watch carefully curates retractions, pulling them from several non-Crossref sources, including PubMed and publisher websites. Each entry is manually checked and annotated before being added to the database. The high level of curation and broad coverage is what made a partnership between Crossref and Retraction Watch attractive, and our shared goal of making changes to metadata more visible.

“Our goal with the Retraction Watch Database has always been for it to be as useful to as many people as possible, and available from as many sources as possible,” says Ivan Oransky, co-founder of Retraction Watch and executive director of The Center For Scientific Integrity, its parent nonprofit organisation. “Integration with Crossref’s REST API is a huge step in that direction.”

Where can I see the retractions?

If you use a service that collects Crossref metadata, you will start to see the Retraction Watch retractions as they are picked up. To access the data directly, you can find retractions from both Crossref members and Retraction Watch in our REST API, for example with the following request for all retractions:

https://api.crossref.org/v1/works?filter=update-type:retraction

Or for an individual record:

https://api.crossref.org/v1/works/10.1177/17588359231172420

In the results here you will see an update-to field:

"update-to": [
{
"updated": {
"date-parts": [
[2023,4,22]
],
"date-time": "2023-04-22T00:00:00Z",
"timestamp": 1682121600000
},
"DOI": "10.1177/1758835920922055",
"type": "retraction",
"source": "publisher",
"label": "Retraction"
},
{
"updated": {
"date-parts": [
[2023,4,22]
],
"date-time": "2023-04-22T00:00:00Z",
"timestamp": 1682121600000
},
"DOI": 10.1177/17588359231172420",
"type": "retraction",
"source": "retraction-watch",
"label": "Retraction",
"record-id": 44124
}
]

The source field states where the retraction came from. Currently, it can have two values: publisher or retraction-watch. Note that the same retraction may be included multiple times from different sources.

Retraction Watch retractions will remain available on Gitlab in csv format and be updated on working days. The record-id refers to the entry in the csv file with further details, such as the reason for retraction.

There is full documentation available for the Crossref REST API and if you are new to REST APIs, see our learning hub to get started which includes a tutorial about accessing retractions.

What can I do with the retractions?

Like the rest of our metadata, the retractions are freely available. If you use or operate a tool that ingests retractions, the new entries will start to be picked up immediately. The Retraction Watch database includes a larger number of retractions than the Crossref database, so you should see an increase in the total.

We have heard from organisations that would like to build new research integrity tools based on this data. We look forward to seeing the benefits brought by wider availability of the Retraction Watch retractions, and how they can provide better context to research outputs.

While Crossref metadata is freely available to reuse without a license, if you make use of the Retraction Watch retraction metadata in a published work, we kindly request that you provide a citation to the source.

If you have questions or comments, please head over to the section of our forum dedicated to integrity of the scholarly record.

News: Crossref and Retraction Watch

Ginny Hendricks — Tue, 12 Sep 2023 00:00:00 +0000

Crossref acquires Retraction Watch data and opens it for the scientific community

Agreement to combine and publicly distribute data about tens of thousands of retracted research papers, and grow the service together

12th September 2023—The Center for Scientific Integrity, the organisation behind the Retraction Watch blog and database, and Crossref, the global infrastructure underpinning research communications, both not-for-profits, announced today that the Retraction Watch database has been acquired by Crossref and made a public resource. An agreement between the two organisations will allow Retraction Watch to keep the data populated on an ongoing basis and always open, alongside publishers registering their retraction notices directly with Crossref.

Both organisations have a shared mission to make it easier to assess the trustworthiness of scholarly outputs. Retractions are an important part of science and scholarship regulating themselves and are a sign that academic publishing is doing its job. But there are more journals and papers than ever, so identifying and tracking retracted papers has become much harder for publishers and readers. That, in turn, makes it difficult for readers and authors to know whether they are reading or citing work that has been retracted. Combining efforts to create the largest single open-source database of retractions reduces duplication, making it more efficient, transparent, and accessible for all.

Product Director Rachael Lammey says, “Crossref is focused on documenting and clarifying the scholarly record in an open and scalable form. For a decade, our members have been recording corrections and retractions through our infrastructure, and incorporating the Crossmark button to alert readers. Collaborating with Retraction Watch augments publisher efforts by filling in critical gaps in our coverage, helps the downstream services that rely on high-quality, open data about retractions, and ultimately directly benefits the research community.”

The Center for Scientific Integrity and the Retraction Watch blog will remain separate from Crossref and will continue their journalistic work investigating retractions and related issues; the agreement with Crossref is confined to the database only and Crossref itself remains a neutral facilitator in efforts to assess the quality of scientific works. Both organisations consider publishers to be the primary stewards of the scholarly record and they are encouraged to continue to add retractions to their Crossref metadata as a priority.

“Retraction Watch has always worked to make our highly comprehensive and accurate retraction data available to as many people as possible. We are deeply grateful to the foundations, individuals, and members of the publishing services industry who have supported our efforts and laid the groundwork for this development,” said Ivan Oransky, executive director of the Center for Scientific Integrity and co-founder of Retraction Watch. “This agreement means that the Retraction Watch Database has sustainable funding to allow its work to continue and improve.”

Please join Crossref and Retraction Watch leadership, among other special guests, for a community call on 27th September at 1 p.m. UTC to discuss this new development in the pursuit of research integrity.

Supporting details

Crossref retractions number 14k, and the Retraction Watch database currently numbers 43k. There is some overlap, making a total of around 50k retractions.
The full dataset has been released through Crossref’s Labs API, initially as a .csv file to download directly: https://api.labs.crossref.org/data/retractionwatch?name@email.org (add your ‘mailto’). Edit: 2024-10-10: The full dataset is available in a git repository at https://gitlab.com/crossref/retraction-watch-data.
The Crossref Labs API also displays information about retractions in the /works/ route when metadata is available, such as https://api.labs.crossref.org/works/10.2147/CMAR.S324920?name@email.org (add your ‘mailto’). If you don’t have a .json viewer, please see below for screenshot.
Crossref is paying an initial acquisition fee of USD $175,000 and will pay Retraction Watch USD $120,000 each year, increasing by 5% each year.
The initial term of the contract is five years. ~~The full text of the contract will be made public in the coming fortnight.~~ EDIT 2023-09-26: Here is the signed agreement.
There will be a community call on 27th September at 1 p.m. UTC (your time zone here). Please register.
An open FAQ document is available to collect questions to be answered at the webinar.
This announcement will always be accessible via Crossref DOI https://doi.org/10.13003/c23rw1d9; please use this persistent link for sharing.

About Retraction Watch and The Center for Scientific Integrity

The Center for Scientific Integrity is a U.S. 501(c)3 non-profit whose mission is to promote transparency and integrity in science and scientific publishing, and to disseminate best practices and increase efficiency in science. In addition to maintaining and curating the Retraction Watch Database, the Center is the home of Retraction Watch, a blog founded in 2010 that reports on scholarly retractions and related issues in research integrity.

About Crossref

Crossref is a global community infrastructure that makes all kinds of research objects easy to find, assess, and reuse through a number of services critical to research communications, including an open metadata API that sees over 2 billion queries every month. Crossref’s >19,000 members come from 151 countries and are predominantly university-based. Their ~150 million DOI records contribute to the collective vision of a rich and reusable open network of relationships connecting research organisations, people, things, and actions; a scholarly record that the global community can build on forever, for the benefit of society.

Enquiries

For Retraction Watch/Center for Scientific Integrity: Ivan Oransky, ivan@retractionwatch.com
For Crossref: Ginny Hendricks, ginny@crossref.org

A screenshot of an example Labs API metadata record with a Retraction Watch-asserted retraction

Towards a connected and dynamic scholarly record of updates, corrections, and retractions

Ginny Hendricks — Sat, 17 Sep 2022 00:00:00 +0000

Updates, corrections, and retractions as part of a connected record

The scholarly record is not static: works are revised, corrected, and sometimes retracted, and readers need to see those changes wherever they meet a work, not only on the version they happen to land on. This 2022 preprint sets out how richer relationship metadata can make that possible.

About the paper

The paper examines Crossref’s role in maintaining trust in the scholarly record, and how corrections, retractions, and errata are expressed in metadata today. It reframes Crossmark: not only a badge to display a work’s status to readers, but relationship metadata that connects a work to its later updates, so that a correction or retraction travels with the work across the many places it is cited, indexed, and read.

That framing places updates, corrections, and retractions inside the wider Research Nexus: the value lies in the relationships between outputs, not in any single record or metadata element in isolation.

The paper looks at where today’s approach falls short, and at how a more connected and dynamic model of the record might develop.

Read the paper

Read the preprint on MetaArXiv

Encouraging even greater reporting of corrections and retractions

Kirsty Meddings — Mon, 30 Mar 2020 00:00:00 +0000

TL;DR: We no longer charge fees for members to participate in Crossmark, and we encourage all our members to register metadata about corrections and retractions - even if you can’t yet add the Crossmark button and pop-up box to your landing pages or PDFs.

–

Research doesn’t stand still; even after publication, articles can be updated with supplementary data or corrections. When research outputs are is changed in this way the publisher should report and link it, so that those accessing and citing the content know if it’s been updated, corrected or even retracted. This also emphasizes the member’s commitment to the ongoing stewardship of research outputs.

Many people find and store articles to read later, either as PDFs on their laptop or on one of any number of reference management systems - when they come back to read and cite these articles, possibly many months later, they want to know if the version they have is current or not.

Removing Crossmark fees

To encourage even wider adoption of Crossmark, and to promote best practice around better reporting of corrections and retractions, we will no longer be charging additional fees for our Crossmark service. This change applies to all Crossmark metadata registered from 1 January 2020. All members are now encouraged to add Crossmark metadata and add the Crossmark button and pop-up box to their publications - and you can do so as part of your regular content registration.

Richer metadata gives important context

We know that there are many more corrections and retractions that are not yet being registered, and to address this, we are now asking all of our members to start registering metadata for significant updates to your publications, even if you don’t implement the Crossmark button and pop-up box on your content. Remember, anyone can access the Crossmark metadata through our public REST API, and start using it straight away - even if you’re not ready to implement the Crossmark button.

Check out how to get started; if you only want to deposit metadata, follow steps one through four. If you also want to add the Crossmark button and pop-up box to your web pages/PDFs so that readers can easily see when content has changed, then also follow the rest of the steps.

Crossmark

We launched Crossmark in 2012 to raise awareness of these critical changes, by asking Crossref members to:

record such updates in your metadata, either as part of your regular Crossref metadata deposit, or deposited as stand-alone data for back-year records
help readers find out about the changes by placing a Crossmark button and pop-up box (which is consistent across all members making it recognizable to readers) on your landing pages and in PDFs

Members can also use Crossmark to register additional metadata about content, giving further context and background for the reader. These metadata appear in the “More Information” section of the Crossmark box. 7 million DOIs have some additional metadata, the most common being copyright statements, publication history, and peer review methods.

Anyone can access the Crossmark metadata through our public REST API, providing a myriad of opportunities for integration with other systems, and analysis of changes to the scholarly record.

Who has implemented Crossmark?

440 Crossref members have implemented Crossmark to date. 11.4 million DOIs have some Crossmark metadata.

	Total DOIs	DOIs with Crossmark metadata	%
Journal articles	80,862,460	10,155,340	12.56%
Book chapters	14,040,646	792,953	5.65%
Conference Papers	6,175,733	457,237	7.40%
Datasets	1,862,852	19,206	1.03%
Books	753,298	239	0.03%
Monographs	469,333	23	0.00%

Of those, about 130,000 contain an update:

You can see which members or journals have implemented Crossmark by viewing the relevant Crossref Participation Report.