Intern talk:Publication types

Copy&Paste from the current version (05/01/24) to revise the article based on it:

This is an overview of publication types (data sets and collections), their dependencies on manuscripts and best practices of their use. For the public summary see Publication types.

Important notes

 * Do not create any more collections because you or the user just want to have a single DOI for a paper. (*)
 * Relation “suppl. to (dependent)” is outdated and should not be used anymore.
 * Please use “supplement to” (independent), but only for those data that have been prepared for the sole purpose of writing a paper or thesis - there are in fact only a few cases where this applies! We will try to find a better solution with DataCite, crossref etc. for the future.
 * Collections with independent childs must be without login, because no data download possible.

Types of data publications
(*) Initially, this was best practice but in the meantime, it is not necessary any longer as technological settings have changed and our cultural change has now to follow. Journals like ESSD or Scientific Data have agreed that this is Best Practice. The paper reference can be attached to all data sets individually using the relation “related to” or “supplement to”. Several data sets can have the now relational type “suppl. to (dependent)” to the same paper. Thus grouping of data sets for one paper using a parent should only take place if there have similar identical content (e.g. CTD casts).

(**) Please do not start the child data set titles with the words “Table X” or “Fig.X” - if necessary put this information in brackets at the end. Technical solutions are in preparation for sorting the order of child data sets in the future.

(***) It is planned to implement a basket to download selected datasets as well as to create a corresponding citation list.

Background information

 * Google spreadsheet by Michael
 * Most recent slides by Michael
 * Crossref schemata
 * ISO 690:2010
 * ISO 690

The public simplified version is under Data publication

Dana's notes from previous discussion with Uwe (2021-07)
"Bibliography" and "editorial publication" are collections in which childs are "independent", as opposed to "bundled publication" or "publication series", where childs are dependent on the parent. They are technically lists of references, theoretically citable, but such citations are not really recommended, as opposed to citations of individual datasets or "publication series" and "bundled publication".

"Bibliography" is the lowest category of “publication”, the scientific added value is very limited. Bibliographies are actually not really intended to be “cited” because they are just a loose compilation of other citations. Uwe explains a bibliography as an example: ''You have written a paper and want to add a footnote, for example, that there is a collection (without guarantee) “Musterman at al collected some datasets about something during the MOSAIC project, available here: https: //doi.org/10.1595 / ... ". The DOI is not a quality feature here, just a persistent link.''

An "editorial publication" is a different case: Here the editors are responsible for the compilation of the data sets and it is also assumed that they have carried out quality checks on the compilation (selection of the data sets, the quality control, thoughts on the comparability of the methods, ...). What is quoted, however, is the compilation, not the individual data records behind the collection. Therefore, only the editor gets credit, and not for the content of the data, but only for his work that he has compiled it. If you want to pay attention to the scientific H-Index and want to give credit to all authors of the underlying data sets, you have to cite the individual data sets - there is no way around it.

How acknowledging editorial contributions is actually handled in the case of a citation index is unknown to us (not 100% transparent). We as publisher provide the information: "XY, as an editor, selected data sets from a subject area, checked them, compared them and wrote something about it" (Editorial) vs. "XY collected data XY himself and made these * own * data sets available as a publication series" (Series ). In the latter case, this is certainly something else in the sense of a citation index, because the work and role of the contributors is different.

"Bibliography": is not included in our wiki, mainly because of it's limitations regarding the citability.

Example of "Bibliography": https://doi.pangaea.de/10.1594/PANGAEA.932295