Date differences

sciwonder

I've noticed that, on occasion, there are slight differences in the dates when importing from PubMed (by entering a PMID) compared to when importing via a DOI link. For example, the PubMed import might say June 30, 2020, whereas the DOI import will say July 2020. Sometimes the difference is a bit more drastic—a difference of more than a few months—but usually never a difference in the year.

This isn't an issue with Zotero, I think, and it's not likely to pose a problem when citing, unless I use a citation style that requires both year and month in the date element, but I had wondered if anyone knew *why* these differences exist in the first place and which bibliographic source would be best to import from.

adamsmith

The reason for these is most likely the distinction between actual publication dates (aka "online early") and official publication dates (i.e. the publication date of the journal issue the article was included in).
I'm not sure how Pubmed and CrossRef (i.e. DOIs) differ in their practice on these -- but if you're looking at recent articles it's possible that one of the two simply hasn't been updated yet and is still only including the online early date. Both of these are excellent sources for metadata. Pubmed is a bit more standardized, e.g. all articles are sentence cased and you have abstracts and MESH keywords, so when available, I'd probably tend towards that, but I find it hard to predict which is going to be better on the dates.

sciwonder

Apart from Vancouver, are there many other styles which require inclusion of month information in a reference list entry for a journal article? I imagine (or hope, rather) that the number is quite small because it seems that date information (apart from year) might be a bit unreliable. I've found a couple of articles that are a few years old now but still differ with respect to the month depending on where they're imported from, with the article itself not including that info.