33

I am a grad student in computer science. Having read many research papers and having worked with an adviser has made me much more aware about best practices concerning presentation and reporting of results in published research.

I have also published a paper when I was a junior in college. Looking back at it now (its been around 2-3 years), compared to the standards set by papers I have read, my work is very sloppy.

More specifically, I have typos in one or two places and an equation is incorrect (I maximize for x when I should be maximizing for x and y). Also, my evaluation is based on very little data (though I do point out that it is very preliminary and further evaluation is needed) and the metrics I have used to evaluate the described technique do not adequately demonstrate that my method is correct or better performing. The evaluation was added after the reviewer comments and therefore was not checked.

Typos aside, as an undergrad, I was ignorant about best practices when it came to reporting results in that specific area. This is partly because me and a friend of mine wrote the paper over the summer without any involvement from our university or professors (I am not making excuses and I do hold myself accountable for these mistakes).

The paper is rather embarrassing since it does not really represent a significant step forward and is presented very sloppily. I love research (I'm planning a career in research) and I feel that it does not accurately reflect my current research rigor. I don't like it showing up in search results associated with my name.

What would you suggest is the best way to deal with this situation? I'm most worried (been stressing about retractions and the like...) by my lack of sufficient data and invalid metrics, as well as the possibility of future employers seeing it.

Note: I just want to point out that I do not believe that the core idea of the paper is wrong, but the way it is presented along with its evaluation. It's also been cited a few times (I fully understand that this is no guarantee of validity).

  • 30
    So you wrote a bad paper before you received any training in writing papers. That's not a big deal. Just move on, nobody will care about the undergrad paper. – Marc Claesen Aug 15 '15 at 10:57
  • 2
    "What would you suggest is the best way to deal with this situation?": Forget about it and move on to new research, your future won't depend on this. – Massimo Ortolano Aug 15 '15 at 10:57
  • 3
    Publish a follow up in which you briefly acknowledge the deficiencies and take the work itself to the next stage, if there is one. – user207421 Aug 15 '15 at 13:03
  • 7
    I strongly disagree with @MassimoOrtolano. Don't forget about it. Learn from it. – JeffE Aug 15 '15 at 22:11
  • 1
    Thank for the comments! It's made me much more careful about what I write in a paper and how I write it. Also, I will forget about it in the sense that I wont keep worrying about it :). – Harry Aug 16 '15 at 6:48
55

Just don't worry about it, and keep doing your current good work.

It's pretty cool that you were able to publish a paper at all as an undergraduate. Anybody who notices your old paper will primarily notice that. The sloppiness will reflect much more on your senior co-authors (who should have been expected to know better, and educate you likewise) than yourself. And if you didn't have much more experienced co-authors, so much more cool and understandable.

Do good work now, and the sloppiness of your undergraduate paper will simply fade into insignificance.

  • 4
    How could one notice that the paper was written as undergraduate? That's not often written on the paper itself and not all readers will check the OP's CV? – Taladris Aug 16 '15 at 4:02
  • 4
    @Taladris The concern is that it will reflect badly on the author. In a context where a reputation of the author's body of work is being formed, it seems likely that the reader would be considering multiple of the author's papers, in which case it's obvious that this is the earliest. Even non-undergrad first papers can be expected to be of lower quality. – Michael Ekstrand Aug 16 '15 at 12:07
20

First of all, your situation is not so unusual: in fields where student-only publications are common (like mathematics and some branches of CS), a lot of academics have one or two publications from their undergraduate days. These publications are virtually guaranteed to be significantly worse than all of their other publications. If you don't believe it, consider the logically equivalent form: "It is virtually guaranteed that the time you spend in a graduate program will significantly improve the research you do and publish."

I think a lot of people are vaguely embarrassed by their undergraduate publications: e.g. in mathematics most undergrad-only publications are in unusually undergrad-accessible fields like graph theory. But then a majority of pure mathematicians go on to study and work in much fancier fields and even (unfortunately, to be sure) to look down upon these areas. But although undergraduate publications can be embarrassing compared to later publications, they are impressive when compared to other undergraduates: a student-only undergraduate publication puts you at the head of the class.

For your paper in particular you sound more embarrassed by some slipshod details than by the work as a whole. It also sounds like you received some subpar refereeing and editorial work: in particular, asking for extra content and then not looking at the extra content is lazy. It can be surprising sometimes what apparently (and even actually, most of the time) reputable journals can publish. (You might think that each paper published in a serious journal would at least have a flawlessly grammatical title. You'd be wrong. But I digress...) What's done is done, but there's a lesson here: assuming that whatever someone else lets you publish is going to make you proud is sometimes an oversimplification and sometimes an outright error.

I would say that for the most part you should just work on feeling good rather than bad about this. The one thing that may be worth actual action on your part is making clear to interested parties that this was an undergraduate publication. There are some obvious clues to this for those who look. In my field (mathematics), up until very recently a majority of graduate students did not publish papers, so it's rather common to see gaps of approximately five years between a mathematician's first publication and all later publications. Even a gap of a few years followed by a steady stream of publication creates that impression. You should certainly have a webpage, that webpage should certainly have a CV, and that CV should certainly clearly demarcate your undergraduate years. If you feel strongly enough about it, you could list that paper separately on your webpage / CV / publication list in a category called "Undergraduate Papers".

Afterthought: if you think the flaws in your paper will cause trouble to researchers in the field, perhaps fixing the paper in some way might be in order. You could post an edited version of the paper on your webpage, and if someone cites the paper you could refer them there. But ask a more experienced academic before you do this to get a sense of whether it's really necessary.

  • Curious: is research-level graph theory really undergrad-accessible? Or is it somehow more likely to contain "easy" problems than some other field of math? – mrm Aug 15 '15 at 22:40
  • @mrm: I think graph theory at all levels is much more accessible than most other parts of mathematics. For instance, I believe that the following is one of the most important results in graph theory: en.wikipedia.org/wiki/Robertson%E2%80%93Seymour_theorem. It is understandable to anyone who has taken an undergraduate course on the subject. Which is not to say that this is an "easy" theorem, obviously. – Pete L. Clark Aug 16 '15 at 3:19
  • @PeteL.Clark Thanks for the advice and suggestions! About the flaws - I feel that we explained the main idea reasonably clearly. My main issue was with the evaluation since the results could be misleading if you don't account for the size of the experiment. But we clearly state that our evaluation is preliminary, the higher numbers are probably due to the restricted scope of the experiment and that a more rigorous evaluation is required. So, I'm thinking that it is ok. – Harry Aug 16 '15 at 7:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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