1

I am thinking of starting a PhD in CS while holding a full time job. If someone has a full-time job and starts a part-time PhD, it's reasonable to expect the PhD to take more than 8 years to be completed. I am wondering how relevant such a PhD would be, considering how fast some areas of CS are advancing.

For example my area of interest (and where I have practical experience) is computer vision and statistical learning. If I had started a PhD thesis, let's say in 2010, then by 2019 my thesis topic would most likely be redundant, considering all the advancement in computer vision and machine learning in the past decade.

I guess I am trying to get more insight from the more experienced people.

4

You don't wait with publishing till you finished your PhD, especially in such a fast moving field. So by the end of your education you will have made a contribution, even if at the time your PhD is finished your contribution is already less relevant. This is fine. A PhD is supposed to show that you can do independent research; think of it as vocational training for researchers. So if you show that you made contributions that at the time were good, then you have shown what you needed to show. Nobody expects of PhD students to be able to see the future.

  • I am sorry, I don't understand the publishing part. If I have a publishable result then why not publish it as my PhD thesis rather than publishing it as a research paper. – zindarod Jul 9 at 11:37
  • A single result is probably not enough for a thesis. – allo Jul 9 at 11:57
  • 3
    @zindarod, there's a difference between a thesis and peer-reviewed papers. The thesis is a requirement for obtaining the degree, while papers (broadly) are used to disseminate results to the interested research community. You would likely publish parts of your research in papers during your PhD, while the thesis would describe all of the research you've done. – Emma Jul 9 at 12:24
  • @Emma Ah, thanks for explaining that. I didn't know that part of the research would be published as papers. – zindarod Jul 9 at 12:27
  • 3
    In my experience (at a highly ranked department in the US), a CS PhD thesis is roughly equivalent to 3–5 papers, most of which are published well before the thesis defense. – JeffE Jul 9 at 13:02
2

In the US, if a doctorate takes 8 years, note that it won't be 8 years from selecting a problem to the end of the dissertation. In other words, a doctorate in the US is normally not completely filled by research directly applicable to the dissertation. It may not be research at all until near the end. It is likely to be only a couple of years if you do it only part time and maybe a year or so full time. But insight into your problem can't be scheduled. There is uncertainty in any research, full or part time. But, of course, you will know more when it comes time so finalize the research question.

Also, your advisor would be unlikely to wait around for an 8 year research project to mature.

The program in the US normally includes coursework related to the general area of the (future) research, seminars in which you get close to the research boundary and explore open and possible problems. There is also preparation for comprehensive exams.

On the other hand, working slowly in a "hot" field with a lot of research activity elsewhere means an increased likelihood that your research will get "scooped" before you finish. That can be a major setback, though sometimes the effect can be lessened.

But, doing what you propose will be a lot like having two full time jobs. It is a recipe for burnout if you don't have ways to manage it well.

  • Yes, I am beginning to get the picture. Even if I had enough hours in the day to fulfill the requirements, I don't think I'd have the energy to maintain a full time job and a PhD. I'd probably burn out in a couple of years. – zindarod Jul 9 at 12:58

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.