This is a common pattern in publications, not exclusive to my field (TCS):

An author submits a paper to a conference (as preliminary version). Then, after acceptance to a conference, virtually the same paper (modulo some modifications, etc mostly in length but rarely in any conceptual level) is submitted to a journal, possibly with a different title (but not always). The latter is called full version. In many cases, the two versions are indistinguishable and in most cases epsilon away (for a small epsilon).

At the end, the author has two publications (conference+journal) from practically the same technical result.

Is this considered as one or two publications? (technically of course are two different publications, but how people perceive that?)

From what I have seen, this is generally considered as 2 publications. So, an author with 5 results that follows the above pattern will have 5 conferences + 5 journals = 10 publications. Another author with again 5 results that are submitted only directly to journals (for one reason or the other, for example lack of travel funding from the institution) will have only 5 publications and, subsequently, will be ranked lower than the first author/candidate.

My question is: How do academic people perceive the same publication that has been published in a conference as a short version and, subsequently, in a journal as a long version where both versions contain virtually the same text/results. How can someone evaluate this when sometimes these two publications appear with a different title?

  • 1
    What do you mean by "count" here? "Count" for what purpose? Mar 24 '17 at 11:07
  • @AlexanderWoo I mean exactly what I write in the following paragraph. Technically it's two publications but in scientific terms it's just only 1 result.
    – PsySp
    Mar 24 '17 at 11:11
  • @AlexanderWoo OK, I changed my wording. Hope it's clearer now.
    – PsySp
    Mar 24 '17 at 11:13
  • Are they considered difeerent by whom? Speaking personally, as a theoretical computer scientist, I do not consider them different papers.
    – JeffE
    Mar 24 '17 at 13:34
  • @JeffE I am also in TCS, and not everybody considers them as different papers. (sample, and personal experience, example: "look this candidate has 3 FOCS/STOC AND 3 SICOMP" no matter that the journals are the same as the conference papers)
    – PsySp
    Mar 24 '17 at 13:40

I don't think this question has a universal or near-universal answer. Different people, or even sometimes the same person in different contexts, will count differently if they need to count at all.

More generally, I don't think anyone actually makes decisions based on raw counts of papers, though plenty of people try to justify their decisions to others based on raw counts.

Two versions of the same paper count as two if the person doing the counting wants to make you look better and count as one if the person doing the counting wants to make you look worse.

  • I also think the "justification" of decisions could definitely lead, sadly, to different metrics.
    – PsySp
    Mar 25 '17 at 0:01

Not the answer you're looking for? Browse other questions tagged or ask your own question.