The editors of Journal C have found that 15 of their recent articles have been assigned slightly different DOIs in the Online First and the final issue versions. This arose from administrative problems with the publisher’s production process and has resulted in duplicated articles in both spaces, and there may be other duplicate articles due to reassigning different DOIs.
The editors of Journal C would like the publisher to remove the Online First version as would happen as part of the expected process when an article is compiled into a regular issue. This would retain the regular issue as the definitive version. Articles cited from the Online First area could be aliased and the DOI resolved to the regular issue.
The publisher has proposed that the duplicate articles be withdrawn and a withdrawal notice published in the Online First area; or alternatively that the Online First articles be compiled into a forthcoming issue and then stamped as withdrawn with an accompanying withdrawal notice.
The journal’s editors feel that the publisher’s proposal is not in line with the intended temporary nature of the Online First space. Creating a new issue and then withdrawing the articles is a duplication of effort and will be confusing for readers, as well as risking the reputation of the journal. It will also create confusion with the citation calculation for databases such as Crossref.
Questions for the Forum
- Is it acceptable (and possible from a technical point of view) to remove the articles from the Online First area?
- If they cannot be removed, is it acceptable to compile them into a separate special issue in the online space which would then be withdrawn and accompanied by a withdrawal notice?
The opinion of the Forum was that this is a matter of publication ethics because it relates to the integrity of the scholarly record. The most straightforward solution is simply to redirect the duplicated DOIs (including that of the Online First version) to the final article. This would bring the record into line with what would have happened if the standard process had been followed, and brings no ethical issues. Other solutions are more complex than is necessary and would bring less clarity to readers.