I am simply looking for a rough estimate on how many PhD applicants typically apply having published a peer-reviewed paper (or papers).

Particularly, I am interested in Computer Science (or STEM fields in general).

Also, it would be interesting to know the same percentage for the admitted students.

These questions can likely only be answered by those on (or previously on) admissions committees, but all responses are welcomed!

  • First-author or co-author?
    – mankoff
    Jun 25, 2013 at 0:03
  • First, preferably.
    – HJM
    Jun 25, 2013 at 4:27

4 Answers 4


I'm in a CS department at a mid-ranked school in the US, and have reviewed applications for Ph.D programs in CS for the last 6 years. I didn't compile detailed numbers, but my sense is that the number of candidates with "actual" publications (as opposed to fluff pubs) is of the order of 5%. I suspect this number is higher for the top-ranked school.

  • 5
    Maybe not. The absolute number of applicants with publications is probably higher, but so is the total number of applicants. The more interesting question, in my opinion, is what fraction of admitted applicants have real publications.
    – JeffE
    Jun 25, 2013 at 2:22
  • Can you, as admissions counselors, identify these fluff publications?
    – HJM
    Jun 25, 2013 at 4:31
  • 1
    It's not a perfect method, but usually if you look at where something got published you have an idea. Most of us in CS have a passing familiarity with the top few venues for publication in each sub area, and we always have experts on the admissions committee who can help out. I don't mind that much though, if the student can articulate something about the actual research in their statement.
    – Suresh
    Jun 25, 2013 at 6:50
  • 1
    I'd be more interested in the distinction between a "fluff" publication and a "real" publication, particularly with regards to the quality of the work. In any given field, say CS or physics, I'm not going to know the respectable journals.
    – abnry
    Jun 25, 2013 at 16:08
  • 1
    You just have to read the paper.
    – JeffE
    Jun 25, 2013 at 20:58

In mathematics, at a 10-20 ranked place in mathematics, essentially no grad-program applicants have an real publications.

About 1/3 may have some (as Suresh put it) "fluff-pubs" as spin-offs from summer REU programs. These are not bad things, by any measure, but are more indicative of the socio-economic class of the applicant than their talent or potential. For that matter, it is sometimes quite awkward to explain to novices that their "publication" is a fluff-pub, not real.

Thus, in fact, there is an actual negative to fluff-pubs on an application, since it suggests a possible unfortunate rigidity or over-confidence.

(Once again, in mathematics, if it were possible to do wonderful research in a few weeks over the summer, why does it take 5 years to earn a PhD? There is a misunderstanding... though, yes, it is good to cultivate enthusiasm among talented beginners! Let's just not lie to them.)

  • This is similar to the situation in physics. There I'd say probably 1/3 of applicants have a "publication" as well, often from an REU but sometimes they turned a wrench a few times in an overly generous lab. I'd argue though that such a publication is never a bad thing (unless I'm trying to talk to you about it and it becomes clear you didn't understand your own paper...) but my department seems to care a lot less about undergrad publications than - at least - what I thought they were worth when I was applying.
    – wsc
    Jun 25, 2013 at 3:31
  • Are these "fluff" pubs obvious to admissions committees? And are those with legitimate, non-fluff pubs have at an advantage in comparison?
    – HJM
    Jun 25, 2013 at 4:29
  • Yes, the fluff is pretty obvious, although it's not construed as anything necessarily bad. Non-fluff is quite rare, and would be accompanied by other highly unusual aspects of a file, so it itself is not the issue, I think. Jun 25, 2013 at 13:57
  • 1
    @paulgarrett, I think the last sentence in your 2nd paragraph is backwards. It should instead be something like "disabuse novices...publication is not real, but a fluff-pub."
    – mkennedy
    Jun 25, 2013 at 16:21

In countries where it is common to do a MSc, many PhD applicants have either published papers or prepared/submitted manuscripts, since a MSc would include a research thesis. The level of the publication can vary, and this can also vary by field (experimental projects typically take longer so probability of publication is smaller).


In computer science at a top-ranked US university, I'd estimate that about half of admitted Ph.D. students have a publication while they were an undergraduate.

So, having a (good) publication is really helpful, but not an absolute necessity. What matters most is research potential, i.e., the potential to be a successful researcher. Showing that you have done good research that led to a publication is one powerful way to show that you have good research potential, but there are other ways (e.g., by doing research, getting strong letters of recommendation from folks you have worked with, excelling in academic work, doing independent work).

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .