12

I am writing an article from my master's thesis. It is a polished and much shorter version of it, with equal contents. My master's thesis is unpublished by my university, but available from my GitHub.

I am unsure whether to include some proofs in detail in my article, because too technical and not particularly enlightening. I have received interesting suggestions here. It occurred to me that a different option might be to cite my master thesis for such passages, where they are spelled out in great detail.

I see many advantages:

  • the thesis is already "peer-reviewed" by my thesis supervisor, a respectable researcher in my area
  • it is available online
  • I would make the article more readable, without filling it with too many technicalities, which are however available for the interested reader in the master's thesis
  • I could include other details and content that would otherwise need to be left out, providing a better service to the community

As disadvantages:

  • the reviewers of my article may not consider the thesis as a sufficiently valid source
  • the article would become less self contained (but it is common practice not to repeat things already done elsewhere)
  • I would be referring to proofs and material which although clear, it is not written and polished with a peer-reviewed publication in mind
  • the master's thesis uses a different notation

Is it okay and advisable to cite my master's thesis? Are my concerns valid and are there more?

4
  • 1
    I don't know what the norm is in your field or for the journal(s) you are targeting with your article. However, inasmuch as you say that the article is effectively a version of the thesis, you should consider whether your prior self-publication of the thesis on GitHub presents an issue for publishing the article in a journal. Journals generally require submissions to be original works of authorship, not previously published elsewhere, and an editor conceivably might decide that your article does not meet that standard. Jun 2, 2023 at 15:06
  • If your master thesis is publicly available on github, then it is published. Maybe not "published in a research journal", but definitely "published".
    – Stef
    Jun 3, 2023 at 7:31
  • 1
    @Stef: “Published” in academic settings has a range of meanings, depending on context — “published in a peer-reviewed journal”, “published in any traditional venue [e.g. lightly-edited conference reports]”, “published in any venue that can be expected to remain accessible indefinitely [e.g. major preprint servers]”, “available in any publicly-accessible form [e.g. author’s website]”, … OP’s thesis is certainly published in the last sense, but certainly not in the first.
    – PLL
    Jun 3, 2023 at 13:24
  • @PLL Hello. Yes, the OP's thesis is published. No, the OP's thesis is not published in a peer-reviewed journal. I don't see any ambiguity here. Good day to you.
    – Stef
    Jun 3, 2023 at 14:30

5 Answers 5

16

I am unsure whether to include some proofs in detail in my article, because too technical and not particularly enlightening.

If the results you are proving are original then you should include the proofs, even if they are highly technical and "unenlightening". For original results the most basic function of a proof is to confirm that the result is definitely true. This function is necessary even if the proof itself does not assist in illuminating the result. The usual place for such a proof would be an appendix to the paper. The main reason you would want the proof to be part of the article is that it is a neccessary part of the demonstration of a result at issue, and a outside source on GitHub could change or be removed in the future. If the results you are proving are not original then you have more latitude; in this case you could reasonably cite relevant proofs in other sources, including the proofs in your Masters thesis.

From your linked question, it appears that the theorem at issue here is an adaptation of another published theorem, but your version is not a special case of the latter so it is not already proved. In view of that, I would recommend keeping it as part of your paper. If you are concerned about the technicality and length of the proof, put it in an appendix instead of the body of the paper.

Finally, since you have just finished your Masters thesis, you should consider going back to your former advisors to get their advice. They are already familiar with the material at issue and they will be in a good position to give advice on the best place for the proof.

2
  • 2
    The OP can also publish the thesis on a pre-pint server such as arxiv.org to ensure that it remains available to everyone in perpetuity. This would remove the problem that "an outside source could be changed or removed". Also in many fields and journals a very summarized style of proof might be allowed or even encouraged. The purpose of most "proofs" in journals is to explain the key insights such that an expert in the field can reproduce them with some work. Not to give a computer verifiable rendition of the proof.
    – Kvothe
    Jun 2, 2023 at 11:47
  • 2
    It's also common practice in some fields to publish an extended version of a paper with such proofs on pre-prepint servers if they would exceed the page limits for the paper submission. Jun 2, 2023 at 13:21
7

It occurred to me that a different option might be to cite my master thesis for such passages, where they are spelled out in great detail.

This can be a good option, and is actually very common especially if you are dealing with a page limit on the present article, or if the present article is focused on something else (e.g. application of the ideas, or summary of their impact), such that the formal development would be a distraction.

However, there is one important caveat: in order to publish and cite the master's thesis, you should put it up via a permanent location such as arXiv or a similar service. A GitHub repository is unfortunately not sufficient for this, as GitHub files can later be moved or deleted.

Addressing your specific points: for the "advantages", I agree with all of them, except:

  • it is available online

By academic standards, this isn't entirely true for the reasons above -- the thesis may be available online on June 1, 2023, but may not be available in a decade or a 100 years from now unless it is put on arXiv.

Addressing the disadvantages:

  • the reviewers of my article may not consider the thesis as a sufficiently valid source

This probably isn't a problem; perhaps surprisingly, many good papers cite unpublished or un-peer-reviewed work, including preprints, invited papers, and even personal correspondence.

  • the article would become less self contained

This is true, but could be an acceptable drawback for your use case.

  • I would be referring to proofs and material which although clear, it is not written and polished with a peer-reviewed publication in mind

This is also a valid concern.

  • the master's thesis uses a different notation

This is probably the concern I'd be most worried about. Are you confident about the results you are using? Do they translate correctly to your new setting?

Overall:

Is it okay and advisable to cite my master's thesis?

Yes, it is certainly OK. Whether it is advisable is somewhat of a judgment call, depending on the material. I would encourage you to talk to your masters advisor to get feedback on the present paper structure.

5

You do not need all the thesis, but simply some proof/results. Put everything useful from your theis in one or more Appendices to your paper. Eventually, you can cite your thesis in the introduction or wherever needed as

Results obtained by Leonardo (2022) are presented in Appendix A-B-C

One may wonder why cited results are presented again in the appendix, but then when one reads the full citation being a Thesis (unpublished, but usually thesis can be found after enquiring the university library), the reader will be glad they can check the appendixes,

Appendices can be extremely concise, you need no frame nor introduction, simlpy put what is needed.

4

From the point of view of a reader of your paper, your masters thesis does not yet exist as a peer reviewed publication.

the thesis is already "peer-reviewed" by my thesis supervisor, a respectable researcher in my area

No, it's not; this is not how the peer review process works. For starters, no journal would ask your supervisor to be a reviewer because of close personal/professional connection. And then, accepting a thesis does not mean that it is accepted on publication level. Neither is your supervisor responsible for pointing out issues and making sure they are corrected in the way peer reviewers should.

it is available online

As pointed out by other answers already, GitHub is not good enough for this and arxiv or similar would surely be better.

However, a better option would be that many journals offer to add online supplements to a paper, which will be included in the peer review process. This is a possibility to write a shorter and more readable paper but still to provide the details to the peer review process, and have them accessible online in an "admissible" place.

the article would become less self contained (but it is common practice not to repeat things already done elsewhere)

As your thesis doesn't count as a publication, if you write a paper for publication based on your thesis, material in your thesis doesn't count as "already done elsewhere".

Final remark: You may want to take into account the general publication culture in your field and in the journals you consider. For example, in (pure) mathematics, often more proof detail is expected than in statistics. You may ask your supervisor about this. Also have a look at how other authors did such things.

-1

Go ahead and reference it. Jay Earley certainly referenced his dissertation in his CACM paper; it would be dumb not to. If you can get your department to house it, that would be preferable to github.

https://web.archive.org/web/20040708052627/http://www-2.cs.cmu.edu/afs/cs.cmu.edu/project/cmt-55/lti/Courses/711/Class-notes/p94-earley.pdf

0

You must log in to answer this question.

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