Sunday, October 07, 2007

Is a journal article in the hand worth two in the bush?

Let's say that you have a paper that you've recently presented at a conference. Your plan is to combine this paper with another conference paper, with some additional extension work thrown in, and submit it as a journal article. (This is pretty much standard practice in CS.) And in fact, you've already started working on it. You feel that the work is strong enough that it has a good chance of being published in a pretty good journal, maybe even one of the "big" journals in the field.

Let's say that you were contacted by the organizer of that conference, inviting you to submit an extended version of the conference paper to a different journal. The paper would still undergo the peer review process, so no guarantees that it would be accepted; but you assume that the odds are pretty good that it would be accepted. You are not sure that this other journal is the best fit for this work, however; in fact, the extensions in order to make it fit to the journal may be tricky (or at least not something that you pictured doing with this particular line of research). And then there is the whole dilemma of what do you do with that other paper, the one you were going to combine with this one. (Most likely, the paper would sit around until the next conference paper gets written, which may not be for a while....)

If you were in this situation, what would you do?


Schlupp said...

I'd stick to the original plan, because it seems to fit in better with your research plans. And more work for the 'new' journal probably does away with the advantage from easier acceptance.

Veo Claramente said...

It would also partly depend on how much you want to the goodwill of the conference organizer or whether they would care if you didn't accept their (generous?) offer.

Jenny F. Scientist said...

Oh, I was going to say what Veo said; what do you get out of submitting it to this journal? Besides, it wouldn't be that hard to say that you already have a MS in prep for another journal but will keep theirs in mine.

I'd stick with the original plan unless I got something useful out of plan B.

Anonymous said...

Do that which brings you one step closer to tenure faster.

Kathi Fisler said...

Many of the (CS) conferences I know have gone this "special issue" route in recent years. The journals use this strategy to attract articles; the conferences use it to attract authors. It contributes to a vicious cycle of journal papers becoming less relevant in CS (but that's off point).

I went through this same situation a few years back, and turned down the special issue to go for a top journal submission instead. For me, the benefits of a top journal publication outweighed the easier route to publication in the special issue. This is more than a tenure issue. You talk about perhaps trying to move schools someday. Depending on the level you want to move to, and whether you'd move pre- or post-tenure, publications in top journals may well matter. It may also matter to your letter writers (who hugely influence tenure and moving).

One problem with these "special issue" papers is that they get invited too soon after the conference for much to have changed in how you view the work. I find the best journal articles are often submitted a couple of years after the original work, when the results have settled in and the author understood the problem at new depths. If your goal is to write a high quality journal paper, time helps.

Overall, I believe aiming for the top journal is worth the shot unless the quantity rather than quality of your journal pubs is more important for your career goals (as it is in some situations). I've also found the quality of reviews to be much higher with the top journals, which just helps your research overall.

Good luck,

Jane said...

Thanks for the advice, everyone! I've decided to not contribute to the special issue---because, as many of you asked, I do need to go for the higher-prestige publications right now. (Plus, that other journal article that I'm working on is probably destined for a lower-prestige journal, so I definitely don't need another article of that caliber right now.) Now I just have to figure out how to craft that email to the conference organizer....

Tommy Times said...

In my experience, Deans count papers. Quantity, not quality. Hew to the maxim of the least publishable unit. Some hyperrighteous scholar might call it unethical, but you gotta play the game the way its played, and you gotta eat.

Ewan said...

You might also, in the email you're crafting, offer to instead write a (mini)review of the appropriate field. That way you get to have both; it's working for me a little right now :).

Jane said...

tommy, I do hew to the LPU, but in my subfield I find that it works better for conference papers. (I'm doing fine on that count, and I'm on pace with the journal articles, but I really do need a Quality article at this point to make my tenure case stronger.)

ewan, that's a great idea! I wish I would have read your comment before sending the email. Oh well, next time.

Shriram Krishnamurthi said...

This is coming rather late, but I hope these will be of use in future situations, and of value to others who read this blog and its comments.

The tone of your remarks, and of some of the (less cynical) replies, suggest a misconception of the relative roles of the chair and the author in this process. I'm sure you (Jane) know this, but for general information, here's how this process works. This is from the perspective of an associate professor in CS who has chaired some conferences and received several of these journal invitations. My comments are specifically written from a CS-centric perspective.

The committee that selects papers (program committee, or PC) usually votes on a set of papers that they think are best. A journal agrees to publish this set of papers, as a bundle, in a special issue.

I have lots and lots of thoughts about this idea of voting for best papers, and of expediting them to publication, and none of them are even slightly positive. (Kathi Fisler's comments reflect some of these views.) But setting those aside, it's useful to note that (a) the chair is not doing you a favor, only passing along the PC's views; and (b) if anything, you're doing the journal a favor by providing them with material. (CS journals these days are increasingly desperate for matter to print.)

The right reaction from you is to (a) put the recognition in your vita (and make sure your chair and tenure committee know about it), (b) thank the chair for giving you the good news, and then (c) do what's right as a scientist (and try to make this compatible with doing what's right for your career). Yes, there is a convenience for the journal and for the chair (who usually will also serve as editor of the special issue) if everyone says yes, and delivers on time, and so forth. But that's their problem, not yours. And anyway, various things do go wrong: (a) some invited authors don't deliver on time; (b) some authors think they're a shoo-in and fail to sufficiently augment their paper, resulting in rejection; (c) some reviewers dislike the paper, forcing rejection; (d) some reviewers demand many more edits, which are inconsistent with the limited time in which the special review papers must be accepted; and so on. Only a naive chair thinks this will proceed like clockwork. So you're not really introducing any new uncertainty into the system.

As for ewan's suggestion that you offer to write a review, don't bother. It suggests a sense of obligation to the chair. This is misguided. You owe the chair nothing. You did good work; congratulations! Don't turn what should be a celebration into a guilt-trip.

Bottom line: Don't feel guilty! A short, polite “No thank you” is all you owe the chair. Don't, out of some misguided sense of obligation, take on an additional burden that you don't need and that nobody will recognize!