I enjoy watching birds, but I don’t keep a life list. I don’t keep a life list for anything, really, which might surprise people who know how data-nerdy I am. The exception: the journals I’ve published in. I don’t really know why I track this, but for some reason I find it fun. (To be honest, I’m kind of proud of it and I celebrate each new addition, but I can’t tell you why and I have a sneaking suspicion that I shouldn’t*).
So here’s my list as of today:
My list includes journals that are broad-scope and narrow ones, high-impact and low-impact, society and independent and corporate, open-access and paid-access. There are all kinds of issues there, and maybe I’ll blog about them later.
What interests me most about my list is that it seems long to me. My P = 69 papers (excluding monographs and book chapters) appear in J = 42 journals, giving me a Journal Diversity Indix (JDI) of 0.61 – where the JDI (which I just invented) is simply J/P. The JDI ranges from zero (an infinite number of papers, all in one journal) to 1 (every paper in a different journal)**.
I speculate, without any data at all, that my JDI is large for my field. If so, that’s probably in part because I’ve been more impatient about publication than I should have been – that is, I’ve often gone for the lower-impact sure thing rather than send everything to Evolution or The American Naturalist. More interestingly, I suspect my JDI is large because I’ve done a lot of field-hopping through my career. I may be the only living human to have published in the Journal of Immunology and also in Forest Ecology and Management.
Once upon a time, having a low JDI was probably a Good Thing. In the days when it was harder to search for papers, journal “fit” was probably more important than it is now***, and it might have been more important to place related papers in the same journal. These were the glory days of the numbered series of papers: “Remarks on the Systematics of Gastroplusiidae I”, “Remarks on the Systematics of Gastroplusiidae II”, “Remarks on the Systematics of Gastroplusiidae III”…“Remarks on the Systematics of Gastroplusiidae LXVII”. Now that fewer people find papers by scanning the tables of contents of their favourite journals, and more use text-mining tools like Google Scholar Alerts, a high JDI probably carries a lower cost in apparency of one’s work. It would be interesting to know whether JDIs have increased as a consequence, although one would have to somehow separate this from the confounding effect of proliferation in the number of published journals.
Now it’s you turn (please leave a Reply). Does anyone else pay attention to their journal life list? If you do, how does your JDI compare to mine? What do you think a low or a high JDI implies about your career? Should one aim to keep JDI low or enjoy when it creeps up (as I do)? Should I stop tracking minutia and go back to my day job? OK, maybe don’t answer that last one.
© Stephen Heard August 3, 2017
UPDATE: I’m not the only one who finds this fun. Here, for example, are Andrew Hendry’s data (although not quite comparable because he includes commentary pieces, which
inflate influence his profile). And via Twitter, here are journal life lists from Damien Farine, Nathan Furey, Jeff Clements, Roland Kays, Chris MacQuarrie, Kevin Wood, NK Simons, Julia Koricheva, and Viktor Baranov. Finally, here are two very early-career life lists, and I’m quite jealous of each of them: Matt Grobis and (especially) Heather Penney.
UPDATE #2: In the Replies, Andrew Jackson provides code to automate extraction and analysis of JDI data. Wow.
In case anyone wants to play around with different indices, here are my data as an Excel file.
*^“Is it a new journal you can add to your life list” does not typically appear either in informal advice for how to choose where to send a manuscript or in formal models of optimal submission strategies. There are good reasons for this.
**^There are probably better indices. The JDI doesn’t capture the shape of the J-shaped curve. To see this: 16 papers in one journal and one in each of 4 others gets the same JDI (0.25) as would 4 papers in each of 5 journals. In an attempt to improve my analysis, I calculated a bunch of other indices, including Lloyd’s (1967) index of mean crowding, M*J = Σ(ni(ni-1))/ Σni, where ni is the number of papers in the ith journal. Then common sense and task prioritization regained control and I stopped.
***^Megajournals like PLoS One don’t have “fit” at all. We are collectively running a very large experiment about whether this matters.