Impact factor trends

A few days ago, I received an e-mail message from the Publisher of my favourite physics journal (JLTP), who was pleased to inform me that its Impact Factor (IF) climbed in 2010 to 1.403, seemingly a significant improvement from the 2009 value of 1.074. Obviously, being not only a reader and a contributor, but also a member of the editorial board of JLTP, I was delighted to hear the good news.
As I have expounded already on this blog, I regard raising its IF as crucial, if the JLTP is to retain not just its prestige, but even its long term viability as a scientific publication venue, given the emphasis placed on IF by university administrators, and generally by those individuals who are tasked with reviewing the performance of academic scholars.

However, I should explain more precisely the meaning of the expression “raising the IF”, because it is not obvious what it is.
First off, my curiosity made me want to find out what had happened, in the same 1-year time frame, to the JLTP’s main competitors, and that is when my initial enthusiasm over the rise of JLTP’s IF was curbed. Here are the data (from ISI Web of Knowledge):

Journal IF 2009 IF 2010 Difference
JLTP 1.074 1.403 +0.327
Phys. Rev. B 3.345 3.772 +0.297
J. Phys. CM 1.964 2.332 +0.368
New J. Phys. 3.312 3.849 +0.537
Phys. Rev. Lett. 7.328 7.621 +0.293

It is clear what the problem is, isn’t it ?
As shown by the table, essentially all journals that are “competing” with the JLTP (i.e., they are suitable publication venues for the same articles), have posted comparable IF gains, the only notable exception being the New Journal of Physics which clearly outperformed the competition this year. Interestingly, even the IF of the prestigious Physical Review Letters, which cannot in fairness be regarded as a competitor of the JLTP, went up roughly by the same amount, as if a rigid upward shift of approximately 0.3 had been applied to all journals.

The thing is, the IF is like the potential energy — its value for any individual journal is meaningless per se, only the comparison of IFs of two different journals is significant. In practice, this means that, in spite of the rise of its IF, since the JLTP has made no relative gains compared to the other journals, it does not look any more appealing to authors today than it did one year ago. There continue to be too many alternative journals with higher IF. This has the effect of greatly discouraging authors, especially academic faculty on the tenure stream, from submitting articles to the JLTP.

It is thus interesting to pose the question: how did all journals go up by roughly the same amount ? I am not sure. I do not know whether all Editors have discovered the same “trick”, but if that is the case, the final outcome is merely a washout.
One simple scenario involves greater scrutiny on the part of all Editors, who may have decided to raise the bar on published articles. How does that end up bumping the IF of all journals ?
Say I am the Editor of journal A whose IF is 3, and I reject an article submitted for publication, on the prediction that its number of citations will not be high. For the sake of argument, let us assume that that article is cited only twice after three years. If that article ends up published in journal B, whose IF is 1, the decisions of the Editor of A of rejecting it and of that of B of publishing it, have the net result of raising the IF of both journals.
Clearly, however, B makes no relative gains compared to A. However, without making any relative gains, its position of inferiority with respect to A remains unchanged, and any excitement over the IF increase is ill-placed, and naive.

At the cost of repeating myself, the only two ways in which the Editor of B can in principle narrow the IF gap with A are, in my opinion, the following:
1) By convincing authors to submit their papers to B, instead of A. This is unlikely to happen, though. Some senior authors, who feel very strongly that B should stay afloat and whose careers are sufficiently well-established that they can afford not to worry too much about IF, might do it… in general, though, why would authors submit a manuscript to a journal of lesser IF, if they feel that they can get it published on a more prestigious one ?
2) By convincing authors of manuscripts that were rejected by a third journal C, whose IF is greater than that of A, to submit them to B instead of A. This strategy, in my opinion, has a greater chance of succeeding. If the IF of C is much higher than that of A (and thus than that of B as well), authors whose aspirations to see their work published in C have been thwarted by an unfavourable editorial decision, may not be overly concerned about the IF difference between A and B. They may be persuaded to opt for the latter by a streamlined, expedite publication procedure (possibly one taking advantage of referee reports produced during the review of the manuscript by the Editor of C).

Tags: , ,

5 Responses to “Impact factor trends”

  1. Says:

    I’ve long wondered*: why is the impact fact given to 3 decimal places? It’s not you, Massimo; it’s just they’re always quoted like that. I’m trying to figure out what possible information we learn from that 3rd (or even 2nd) decimal place**. Maybe the mass of the Higgs?

    *ok, an experimental colleague pointed this out to me, so I didn’t originate the quibble.

    **this comment given in the spirit of the title of your previous post.

  2. James Says:

    Do number of citations to an author’s individual papers ever trump the journal in which an author’s papers appear? For example, is it ever “better” to have a PRL which has been cited ten times or a Physica C that has been cited 100 times?

    • Massimo Says:

      James, in my opinion, and in the opinion of many (e.g., Jorge Hirsch), yes. A highly cited paper is important even if it is published on an obscure journal; conversely, a paper that no one cites is irrelevant, even if it is published in Nature (which is why it will lower Nature’s IF). This is exactly the rationale underlying any evaluation criterion based on citations, rather than publication venue. Seeking a high profile publication is, from the scientific standpoint, only justified if by virtue of being published on that journal the paper will be more widely read, and hence presumably more extensively cited. Of course, because some time is required in order to ascertain the impact of a paper, and because most universities insist with evaluating their faculty annually, they have to base their evaluation on the journal, for in twelve months a paper cannot realistically gather that many citations anyway.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: