A Dataset of Metadata of Articles Citing Retracted Articles

This dataset comprises of metada of articles citing retracted publications. Originally, we obtained the DOIs from the Feet of Clay Detector of the Problematic Paper Screener (PPS - FoCD). Additional columns that were not provided in PPS were added using Crossref & Retraction Watch Database (CRxRW) and Dimensions API services. This detector flags publications that cite retracted articles with additional metadata. 

By querying the Dimensions API with the DOIs of the FoC articles, we acquired information such as more detailed document types (editorial, review article, research article), open access status (we only kept open access FoC articles in the dataset since we want to access the full-texts in the future), and research fields (classified according to the Australian and New Zealand Standard Research Classification (ANZSRC) Fields of Research (FoR), comprising of 23 main fields such as biological sciences, education.

To get further information about the cited retracted articles in the dataset, we used the joint release of CRxRW. Using this dataset, we added the retraction reasons and retraction years. 

The original dataset was obtained from the PPS FoCD in December 2023. At this time there were 22558 total articles flagged in FoCD. Using the data filtering feature in PPS, we had a preliminary selection before downloading the first version of the dataset. We applied a filter to obtain:

non-retracted citing articles at the time of data curation*

open-access citing articles since we need the whole text to go forward with natural language processing tasks

cited retracted articles with at least one scientific content related reason of retraction 

only articles (not monographs, chapters) to retain a unified text type

More information about the usage of this dataset will be updated. 

*Current retraction status of the citing articles can be different since this is a static dataset and scientific literature is dynamic.