Science bibiography style incorrectly gives reference as "In press"

dww1309
If I use repository style for the journal Science, the bibliography for this article incorrectly gets formatted as "In Press" when it is not

URL to article:
http://onlinelibrary.wiley.com/doi/10.1002/2015GB005239/abstract

How can I correct this? NB if I use other styles this does not happen.
  • dww1309
    I managed to make some progress on this myself using the graphical editor (http://editor.citationstyles.org/visualEditor/). It seems that if the page number is not a number (in this journal it is an alphanumeric code), then style assumes that it is in press. Deleting this conditional seems to fix it
  • bwiernik
    You shouldn’t have to edit the style. What exactly do you mean “formatted as in press”? Can you post what you are seeing?
  • adamsmith
    The style adds the DOI when page isn't numeric: https://github.com/citation-style-language/styles/blob/master/science.csl#L47

    I agree that's probably not a good decision.
  • dww1309
    @bwiernik - thanks for your comment. This is what the citation style produces

    1. Y. Luo et al., Global Biogeochem. Cycles, in press, doi:10.1002/2015GB005239.

    However, the article is not in press. It should appear with the title, date and page number. Something like this:

    1. Y. Luo et al., Toward more realistic projections of soil carbon dynamics by Earth system models. Global Biogeochem. Cycles 30, 2015GB005239 (2016).
  • dww1309
    @adamsmith Yes I agree that is not a good decision. Also it inserts "in press" when page is not numeric: https://github.com/citation-style-language/styles/blob/ade21027cf1278018047f08af9682fb9af881b44/science.csl#L183
    ...Also not a good decision.
  • bwiernik
    edited January 23, 2018
    @adamsmith I disagree, actually. Is-numeric seems like a good way to test if a page is actually a page number or a DOI stub.

    @dww1309 That article as you cite it IS in press. It hasn’t been assigned an issue or final page number. This is how most publishers do “in press” articles nowadays. So it should be cited using the DOI. “2015GB005239“ is not a page number, but a DOI stub. I think citing using the DOI in this case is exactly what sciencemag wants. Now that the item has been finally published (it has an issue [1] and page number [40-56]), you should edit your Zotero item to include those data.
  • bwiernik
    “in press” seems unnecessary (though the status variable should probably be added for references that are actually as yet unpublished), but the DOI for a paper not yet assigned an issue is definitely correct per the author instructions:
    http://www.sciencemag.org/site/feature/contribinfo/prep/res/refs.xhtml
  • adamsmith
    I don't know -- lots of electronic journals never have regular page ranges -- they're still fully and regularly published articles and the article number can be treated like a page number. (I think citations should always include DOIs, but the journal obviously doesn't agree)
  • bwiernik
    Agreed, but I don't know of any journal that uses non-numeric article numbers. So I think that is-numeric is still a reasonable test for whether the "page/article number" is just a DOI stub from the publisher (stubs that are numeric notwithstanding).

    Also, for any articles that might have a non-numeric article number, Science does want the DOI printed anyway, since they would all be electronic-only journals.

    More generally, I really don't like that translators import the DOI stub into the page field for advance online/online ahead of print articles. I personally wish that data would just be dropped, but I'm not sure if that's the right solution.

    (And, yes, frankly nothing BUT the DOI matters for the citation...)
  • adamsmith
    If we're able to drop the DOI stub on import for print-journals only, I'd be all for that.
  • bwiernik
    edited January 24, 2018
    Publishers are inconsistent in how they report OnlineFirst article pages.

    Wiley leaves the page fields blank in their RIS and use C8, which appears to get imported into page for some reason.
    http://onlinelibrary.wiley.com/doi/10.1002/2017GB005671/full

    SAGE stupidly reports this stub in the page PS field of RIS itself:
    http://journals.sagepub.com/doi/full/10.1177/0149206317744251

    One easy thing to check would be to compare the page value to the stub of the DOI after the publisher prefix and remove the page value if they are the same. I'm not aware of any online-only publishers that use just the article number as the DOI (e.g., Frontiers and MDPI compose a longer stub based on the journal and volume,

    (Aside: the EM translator misses article numbers for Frontiers. Frontiers uses citation_pages for its article numbers, which isn't considered in the EM translator.)
    https://www.frontiersin.org/articles/10.3389/feduc.2017.00002/full
  • dww1309
    edited January 24, 2018
    @bwiernik - thank you. Yes, I see now that if I click on "view citation" on the journal website, there is indeed as you say a numeric page range assigned (which I can edit manually into the zotero metadata). However, in that case it seems maybe there's a problem with the translator for this journal, because if I click on save to zotero, the page field still gets filled with the doi stub. I'm using zotero connector 5.0.31
  • adamsmith
    Yeah, something isn't right in the translator -- looks correct on Wiley's end so this seems to be our problem. I'll take a look
  • DWL-SDCA
    @bwiernik ...but I don't know of any journal that uses non-numeric article numbers...

    Actually, quite a few publishers add an "e" at the front of the article number (some use an upper case "E"). Other publishers use other characters. If it is really needed I can begin a list of these publishers and journals.

    Often, but far from always, the numeric part of the article number is included as part of the DOI.

    Another consideration is a preference by faculty at several universities in California (& I guess elsewhere) to add an "e" preceeding an article/item number so that a reader will know that the article's length isn't limited to a single page.
  • DWL-SDCA
    Here are 2 publishers that precede page numbers or article locators with an alphabetical character:

    https://journal.cpha.ca/index.php/cjph/issue/view/333

    http://journals.plos.org/plosone/?section=recent

    I'll add more to the list if I'm told this might be helpful. I will add at least one more example of a publisher that uses a character other than "e".

  • bwiernik
    edited January 24, 2018
    Probably not necessary to list more.
  • haroldmao
    is there a way to correct this problem? Any update on science style?
  • damnation
    What specific citation do you have the problem with. Can you share a link or the DOI?
  • haroldmao
    the journal is GRL, like this
    Hirschberg J., McArdell B. W., Bennett G. L. and Molnar P. (2022) Numerical Investigation of Sediment-Yield Underestimation in Supply-Limited Mountain Basins With Short Records. Geophys. Res. Lett. 49, e2021GL096440.
    page is " e2021GL096440", that will present "in press"
  • adamsmith
    I'll take this one
Sign In or Register to comment.