I tried asking this question in cstheory.stackexchange.com but it was closed and it seems like this question is more appropriate here.

I am an undergraduate and an American citizen who recently applied to Computer Science PhD programs in the US. Based on conversations with friends and posts on gradcafe I am very likely to be rejected from all the programs I applied to (I haven't heard anything while others have gotten acceptances and interview requests from all the schools I have applied to). I am now trying to brainstorm ideas for what to do after my graduation in order to improve my chances of getting into a PhD program when I try to apply again next year.

One option that seems to be brought up a lot is to attend a master's program. However, unless I can get funding, or transfer credit will lower my tuition significantly (I'll have 6 or 7 graduate level cs courses that will not count towards graduation requirements by the time I graduate), I'm not sure if I can afford such a program.

Another popular option is to be a lab technician or something similar. I'm not sure if such positions are available in theoretical computer science.

Is there anything else I can do?

  • Once you get the official rejection letter, try to figure out what made your application weaker than the others. If you had contact to someone involved in the admission process before, try to ask personally or on the phone. Rejection letters typically give only very general reasons for rejection, but some people are willing to give more specific advice verbally.
    – silvado
    Feb 4, 2013 at 10:22
  • 1
    just wondering, what makes you think you will be rejected?
    – posdef
    Feb 4, 2013 at 10:57
  • 1
    All of the schools I've applied to have either already sent out acceptances, or have sent out interview requests. I haven't gotten any acceptances or interview requests.
    – user13491
    Feb 4, 2013 at 16:10
  • 1
    It seems that your question is US specific, could you please edit it and add more context to it? I guess, the answers might differ depending on the country.
    – walkmanyi
    Feb 7, 2013 at 8:42

5 Answers 5


It's difficult to really answer your question without actually seeing your application, but here's some general advice.

  • Remember that the admissions process is random. — There is nothing you can do to absolutely guarantee admission anywhere. The most you can do is maximize your expected return.

  • Calibrate your expectations. — Are you really a good candidate for a top-5 department? (Hint: Do you have a STOC/FOCS/SODA paper?) For a top-10 department? For a top-25 department? Really? Be respectful but brutally honest with yourself. Ask your letter-writers or other faculty mentors to be brutally honest with you as well. Listen to them.

  • Identify potential advisors. — Every department you apply to should have at least two faculty, preferably more, whose specific research interests closely match yours. Your research statement should not only name those faculty but explain why you think they'd be a good match. Ask your references (or other mentors) for feedback and advice. Listen to them.

  • Spread your applications. — The rule of thumb I heard when I was applying was apply for four schools where you have a reasonable chance of being admitted, one or two backup schools, and one dream school. Do not let the backup schools know that they are backup schools!

  • Get good letters. — Your letters must address your potential for research in personal, specific, and credible terms. A letter that only describes your performance in class is worthless. Your letters must come from research faculty — not PhD students, not postdocs, not lecturers, not managers. If possible, your references should have direct experience with strong PhD students (either as a reference or as an advisor) to make direct comparisons. If possible, your references should be well-known active researchers, but this is actually less important than experience with students. Since you've taken half a dozen graduate classes, you should be in good shape here.

  • Write a good statement. — Your research statement (or "statement of purpose" as everyone bizarrely insists on calling it) must address your potential for research in specific and credible terms. Do not start with an inspiring quotation. Do not write about how computers are changing the world. Do not write about how you've been programming since you were in the womb; nobody cares. Write about your research. Describe your experience. Describe your specific interests (not just "theoretical computer science"). Describe a problem that you might want to work on, with enough background and technical language to convince the reader you know what you're talking about and that you actually care. Bonus points if you correctly cite one of your potential advisor's recent papers, but don't force it.

  • Get feedback. — Send the final version of your research statement to your letter-writers (or other faculty mentors) and ask for their brutally honest feedback. Give them plenty of time. Expect to get your statement back soaked in red ink. Expect different people to give you conflicting advice. Listen to all of them. Lather, revise, repeat.

  • But this is all about the form of the application. The best way to improve the content of your application is DO RESEARCH. Get paid to do research if you can, but do research anyway if you can't. Find a mentor (at your undergrad institution?) if you can, but do research anyway if you can't. Post technical questions and answers to cstheory.stackexchange. Follow CS theory blogs and read the papers that they write about. Keep a copy of the most recent STOC or SODA proceedings nearby to read while you're compiling, or waiting for the laundry, or riding the bus, or whatever. Talk regularly with your letter-writers about your progress. Write, write, write.

  • 1
    Random? Is there another definition of that word? Because, as I know it, random means that its outcome is completely unpredictable, and dictated solely by something akin to a dice roll, as opposite to decisions or algorythms.
    – o0'.
    May 7, 2015 at 13:26
  • 2
    @Lohoris Yes, I mean the standard dictionary definition of random. Running the same process with the same people and the same applications will yield different—and unpredictably different—results.
    – JeffE
    May 8, 2015 at 1:21
  • What is STOC/FOCS/SODA paper?
    – motiur
    May 25, 2015 at 3:52
  • @mothur STOC, FOCS, and SODA are the three most prestigious conferences in theoretical computer science.
    – JeffE
    May 25, 2015 at 16:01

If you intend to do a PhD, then you're tackling the academic career path, so you're at the right place here :)

I am as well applying for a PhD currently, and went through the process of applying for a masters, here're the things I learned the hard way, In general, there are two phases in acceptance, the university acceptance and the departmental acceptance

  1. You should avoid applying to programs directly without contacting anyone in the department you want to work in. So first, check the faculty's members, see their research interests, and contact one of them, telling him that you would want to work under his supervision for your masters because you find his research topics matching your interests and expertise. That person may then make the departmental admission easier.
  2. Most of us google for the top most universities and apply there, which makes our chances lower. Try to look for the 100-200s or search for new offers.
  3. When you contact faculty members, NEVER COPY PASTE EMAILS! these will be very easy to detect, and will result in considering you a spammer. Make sure you tailor your email on the person's research interests and write your most interesting qualifications in the body of the email (not as an attachment, because usually they are too busy to check that, unless you really impressed them through the email's body)
  4. Try to target funding organizations that give out scholarships based on minorities, ethnic groups etc.. these are easier to get accepted in, than the ones available for all the public
  5. Narrow down your focus of the research area that you like most and would love to work in. If you have worked in that area before, it will make your application more unique.

So take care of the above points the next time you apply. As for the skills you can work on

  1. Do some research, try to publish scientific papers
  2. Work on a research proposal, learn how to write a good one as it will help you in future applications and when contacting faculty members. Try to contact your undergraduate university's professors and try to join a research group or work with one of them on a topic that you can publish papers from.
  3. Take the GRE General test, that is a must in most of the universities in the US
  4. Take the GRE computer science test, which is optional in most of the applications but distinguishes your application

If you are not from an English speaking country, make sure you

  1. Take the TOEFL exam, most programs require 80/120, in US they usually ask for 90, which is pretty easy for you from what I can see in your question (take care as the score expires in 2 years, it has to be valid till the time of being admitted, not just the application time)
  2. Make sure all certificates/transcripts are translated to English (by a trusted entity)

    It is usually difficult to get funding for your Masters, because it is mostly studying and not actual working, and you're just staying for two years, thus not contributing that much to the funding organization, but keep trying, never underestimate yourself and go on applying. And remember that even if you keep failing, you are still a long way ahead of those who never tried.

All the best

  • 1
    Take the GRE computer science test — If you're going to do that, you'd better do it soon!
    – JeffE
    Feb 8, 2013 at 3:37
  • 4
    You should avoid applying to programs directly without contacting anyone in the department you want to work in. - In my opinion, this is poor advice. I know it seems to be widely believed among some student populations, but it's more false than true. Please do not spread this meme. A random email that you spend 10 minutes on does not help your case. See my detailed explanation elsewhere.
    – D.W.
    Feb 8, 2013 at 6:55
  • 1
    I still maintain that emailing faculty members is important. If you manage to impress an internal professor, and exchange more emails and discussions, he will definitely have a positive influence on the departmental admission (which is usually the second phase of admission at any decent place). Copy/pasting emails is a bad idea as I said, you should pick the professor with the most relevant research work, with whom you can talk and brag about your achievements. So probably you'll end up finding at most 2 professors that match this criteria in a department that researches in your area of focus. Feb 8, 2013 at 12:20

The most important thing you can do to strengthen your application is: get additional research experience. PhD programs focuses on research. This means that the most important criteria for admission is arguably: likelihood for success at research.

One of the best ways to demonstrate the likelihood that you will be successful at research is to provide evidence that you've already been successful at research. To do that, you need to get involved in an active research and do some serious research. So, my recommendation is: go do some research. If you've already done some, do some more. The more successful research experience you have, the better your odds of being admitted in the future.

Beyond this, it's hard to give more specific advice without understanding why you were rejected and what were the weakest aspects of your application. Therefore, my recommendation is: contact a mentor you trust (a faculty member who is active in research at a Ph.D. program) and ask them to review your application and give you advice about how to strengthen your application.

  • How would a college graduate go about getting research experience?
    – user13491
    Feb 8, 2013 at 15:52
  • @user13491, that's a different question! I suggest posting it separately; you'll probably get some good suggestions.
    – D.W.
    Feb 8, 2013 at 17:06

How are your grades? Grades are a large factor when it comes to Ph.D admissions and it may be worth taking courses that enable you to increase your grades.

Presumably, you are applying to a University abroad, but if there is a local University that performs research where you are, then you may want to work with a professor at your institution for a summer. The chances of getting this type of position may be slim but if you do get one, it gives you some research experience that you can put on your application materials.

If neither of these are good options, then you will probably want to spend your time reviewing related work in the area and working on your research statement. Write it using principles from the foolproof grant template. As a potential student, you might not use all of the elements (as you are limited in both experience, as well as getting about a page's worth of writing) but you should follow the structure for the "first two paragraphs".

  • My overall GPA is 4.84, and my GPA in just math and cs courses is 4.93 (both are on a 5.0 scale). I think these are inflated however, because I took graduate level courses and my school has liberal add/drop policies. I've also been thinking about trying to do research with a local professor, but I don't think such a position is likely to happen as you said.
    – user13491
    Feb 7, 2013 at 21:33
  • 2
    Grades are a large factor when it comes to Ph.D admissions - In my experience, grades are not a large factor for top PhD programs. Poor grades can get you rejected, but good grades cannot get you accepted. Research experience and strong letters are more important.
    – D.W.
    Feb 8, 2013 at 6:58

I have applied for FALL 13 in US and awaiting results. Therefore what I am writing here is a mix of my own experience and the accumulation of my understanding of others (includes forums/blogs).

I applied to all colleges which I thought were good in theory (my area of interest). This list was basically influenced by -

i) the papers I read during my master/undergrad and

ii) the fact that each of the college in the list must have at least 3 potential advisors in theory.

Now the BIG problem was : am I a good fit for these colleges? Frankly, these colleges never disclose their candidate profiles and mostly, the home-pages of the current PhD students do not exist. So I went to forums like gradcafe where there is ton of data but very little useful information but still worth a visit. Some people suggest to mail prof before you apply. In my case I have been advised to not to contact them unless I have strong reasons. Plus, I do not expect any prof to evaluate my profile and see if I am a good match.

What do the admission committee look into any candidate? - research potential. If you have published work then it speaks for itself. Otherwise we have typically three recommendations and statement of purpose (sop).

For recommendation letter, two things matter -

i) Is your recommender known in your area?

ii) How good he knows you? (your association)

Now ideally you should have done some research work with your recommender.

SOP is important and can be seen as you recommending yourself. It also shows your proficiency of writing and communicating with others.

GRE,TOEFL - Some basic cut-off should be cleared. I do not have any idea about the cut-offs.

Funds - If you can manage funds then you cost nothing to college and therefore would be preferred by them.

  • 2
    It only makes sense to contact a prof if you have a specific questions about their specific research projects. One of my most successful long-term professional relations grew out of email I sent to a prof asking his advice of the application of a method he recently developed. I usually regard emails from potential students as spam; from most of them, it is clearly seen that they have little clue about what I do, as they are basically cliches with keywords taken from my webpage [ctd]
    – StasK
    Feb 8, 2013 at 17:23
  • 2
    A real email goes like, "I stumbled upon your webpage as I have been researching grad programs. I've been banging my head against the wall with [YOUR RESEARCH] for the last couple of weeks. In my application, [WHY MY APPLICATION DOES NOT FIT YOUR RESEARCH EXACTLY]. Is [YOUR METHOD] really applicable, or should I look elsewhere?" That's a much better introduction than "Dear Esteemed Sir/Madam, I know that you are the international expert in [KEYWORD FROM MY PAGE], and I want to come study [THE NAME OF THE DISCIPLINE] under you". But yours must be a real problem, and I can tell if it is.
    – StasK
    Feb 8, 2013 at 17:37
  • 1
    Also, your recommendation letters should say how good you will be in research. If your basic degree is in say astronomy, and you've demonstrated all the good habits of a good researcher (self-motivated, able to find the relevant works in the literature, able to tell a good paper from a lousy one), and researched everything on image processing, so you are now more interested in that than in stars per se, it would still be a very good recommendation even if nobody in the prospective CS department really knows that star plasma physicist you've started working with.
    – StasK
    Feb 8, 2013 at 17:41

You must log in to answer this question.

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