Citing stata

[Joe Tuckles]

Hello, can anyone advise what these citations would be classed as:
https://www.stata.com/support/faqs/r...entation-faqs/

Examples include computer program, dataset, book, journal article, online database, report..

[Leonardo Guizzetti]

If you mean the Stata software, it would be a computer program, or a manual would be closest to a book (I'd suppose), if you need to choose from among a preset list in a citation management software. In biomedical literature, software citations are almost always short changed, not elevating to a full reference which would appear in a reference list, but most often given an in-test citation similar to " ... analyses were conducted in Stata (16.1, StataCorp LLC, College Station, TX)". If you have the ability, a full citation is always preferable.

[Nick Cox]

Authors of community-contributed commands always, in my experience, appreciate citations, or at least acknowledgments. Many (including me) would want to underline that the effort and expertise in developing and maintaining particular commands can easily exceed that behind particular published papers of our own.

The bigger deal in practice is whether such citations are customary in any field (my impression is that they aren't often) or are acceptable to reviewers and editors of your paper. Such citations are easier if there is a paper in (say) the Stata Journal or the Journal of Statistical Software that can be cited. Journals will often sniff at citations to commands if the citation is to some other internet source.

The inequalities go further. Anyone with any experience knows that data management is often very time-consuming but authors wanting to be candid and detailed about the process of getting data in shape -- including the commands they used -- would often be told by reviewers to cut that down (or cut it out). Standards are rising to the extent that having reproducible code for a project is more often an expectation, but no one really wants to read someone else's do-file thousands of lines long. The point may more usually be that if something seemed off there is more scope for checking how it arises. Conversely, it seems to me that very few texts prepare students for the shock of how messy and awkward datasets can be. Perhaps the authors don't want to make the subject seem even more complicated and difficult.