GOA8: Two questions that need feedback

Unlike the longer-term question of whether I should do the country book, these two questions need to be resolved before I start datagathering (on or about January 1), as they’re about gathering data. As always, you can comment here (for two weeks), or send email to waltcrawford@gmail.com, or reply to the tweets I’ll do on @goajwcc.

1. Should I rely on DOAJ for fee data in almost all cases?

That is: rather than going first to a journal’s website, looking around to see what it says about charges, and going to the DOAJ info only if I can’t figure things out directly, should I do it the other way around: Start with DOAJ, and unless the journal has had fee complications in the past (e.g., requiring membership. charging a variable fee, charging for submission or charging for both submission and processing), use the DOAJ data?

For GOA7, I wound up using 470 fees from DOAJ because I couldn’t determine the fee otherwise, and there were about 570 special cases. In those 570-odd cases, I would of course continue to base fees on the website itself.

I’ve seen very few cases where the DOAJ information contradicts what’s on the journal’s website, and relatively few journals seem to add complexity to their fees. If I make this change–which would save a fair amount of time–I’d guess a couple of dozen journals would wind up with slightly less accurate fee information (but that info would necessarily be more consistent with DOAJ). Since most complex cases are also relatively small journals with relatively low fees, I can’t imagine that discrepancies would change overall figures much.

I’m inclined to make this change, but I’m certainly open to your thoughts.

2. Should I look at DOAJ first for article counts?

As things stand, I look at a journal’s website first to do article counts–but if it’s not easy to determine the counts, I go to DOAJ and use that count if there is one and if it seems reasonable.

If I switched that, then for the journals that report metadata to DOAJ at the article level–by no means all of them. (DOAJ no longer reports that count on the homepage: my best guess is that about two-thirds of journals report at the article level.)

I would only use the DOAJ article count for a journal if it seemed to make sense–usually only for journals that have been around at least since 2021, where I can compare the DOAJ count with the GOA7 count. If in doubt, I’d try to count the articles directly.

This could save a lot of time (and as DOAJ grows and I get older and slower, time becomes more of an issue). I’m not sure whether it would decrease the accuracy of the figures–and, again, the figures would necessarily be closer to those in DOAJ.

[When nearly all DOAJ-listed journals provide article-level metadata and simple pricing, I’ll stop doing the GOA series, probably, if it doesn’t stop before then.]

Again, I’m inclined to make this change, but definitely open to persuasion.

2 Responses to “GOA8: Two questions that need feedback”

  1. Jan Erik Frantsvåg says:

    1. I have misgivings about using DOAJ APC figures, in my OA publication fund work I often come across journals where APC data are out of date. Journals obviously don’t see it as their work to keep them updated. I think DOAJ has started processes to improve this, but I am uncertain how far they’ve come. For complex models, I would use DOAJ data as these are probably a result of some communication with the journal to establish a reasonable level.

    2. Article counting is hell, I most definitely understand your wish to find better sources. No solution will be THE solution. DOAJ article counts are fine, if they exist and are current. Many journals either don’t deliver, or deliver data late. Data should probably be OK if date of last update is around the end of the year, or into next year. For DOI-using journals, maybe CrossRef might provide better and more updated numbers through some mechanism? I know data exist, but not how to (easily) find them. A problem with using CrossRef is that some journals use DOIs for other than scholarly articles, I think sorting that out might be a problem. I’ve seen a Norwegian journal with article counts way beyond reality, that uses DOIs for nearly anything that could be read. Could LENS https://www.lens.org/ or other freely available bibliometric databases be something to be explored?

    Hope you find a way through this!

    Jan Erik

  2. Walt Crawford says:

    Re fees (APC and other): My impression is that DOAJ doesn’t break down complex fees, so I’d keep taking those from the site. Not sure there is a good answer here.

    Re article counts: I’d only use DOAJ where it “makes sense”–e.g., there’s more than one year and previous years align with my counts. I can’t see involving CrossRef–I need to save time on each journal, not add new data sources. Ditto with other sources, I’m afraid. But thanks; gives me something to think about.