As the title states, I don't have Computer Science background (B.A. in Business Administration and Master of Social Work) but would like to do PhD in Computational Social Science (Data analysis or Information sci).

There are master programs, mostly professional programs, admit students from whole range of areas, such as Computational Analytics and Public Policy from U of Chicago or QMSS from Columbia.

But PhD programs don't. Although PhD programs state on their website that they accept students regardless of their backgrounds, it means that they accept students from mathematics, statistics, or linguistics, but not from Social Work.

My interest is solving social issues by using computational approach. I have taken MOOC courses, such as Linear Algebra and Statistics with R, and I learned R and Python(both intermediate). I also have been working as RA in ICT in Bangladesh project.

To me, it becomes clearer that what PhD programs want is those who have solid CS knowledge and the idea of doing PhD seems to be so unrealistic.

Is it really impossible to do PhD in Data Sci or Info Sci without direct/indirect CS knowledge? If it is possible what should I do?

  • How much Math and stats courses you took while you study B.A. in Business Administration?
    – Nobody
    Sep 2 '16 at 9:29
  • I did Introduction to Statistics (C+), Introduction to Economics(A-), Principles of Accounting(B-), Management Information Systems (B+), Microeconomics(B0), Information and Society (A+), Mathematics in Society (C+), Management Accounting(B+), Optimal Control Theory (A+), Financial Management (B+).. Total 30 credits
    – Renee
    Sep 2 '16 at 9:57
  • Technically speaking, if you know about stats/machine learning and learn scipy and r, you should be good Sep 2 '16 at 13:00

Based on your question and your comments (this and that), I do think you need to take more math and statistics before you apply for PhD. I don't think MOOC courses, such as Linear Algebra and Statistics with R are enough for you to be considered qualified to get into good Computational Social Science programs (You didn't even taken Calculus).

I would suggest you to apply for an MS in computer science first. Once you receive formal and rigorous math and CS trainings, then you apply for Computational Social Science PhD programs. You will have much better chance to get in first or second tier schools this way.

  • I think you are right. But one of the biggest drawbacks of me applying those master programs stated above is funding. Professional masters usually don't provide generous funding. But if it is academic MS program which could be led to PhD, I think I can find some funding? I suppose? Anyways, now it is good that I know clearly that I should not waste too much time and effort looking at and applying for PhD programs.
    – Renee
    Sep 2 '16 at 13:46
  • @Renee Funding for CS masters program is always hard (even for research/academic oriented programs) because too many applicants. But, I do have a suggestion for you. Try to use your BA in business and MS in Social work to get an research assistantship. Not too many people have the same background as you do. So, you may have better chance there.
    – Nobody
    Sep 2 '16 at 13:57
  • 1
    Thank you very much for your kind encouragement! I have been thinking about it too. Though I have different and somewhat irrelevant background, it may become strength rather than weakness. Of course, may not always, but as you say, it is possible, I believe.
    – Renee
    Sep 2 '16 at 13:59

After going through the admission criteria of a couple of universities, what I can summarize about eligibility for PhD in Computational Social Science is the following (they nearly all state the exactly the same points).

  • Degree in any one of CS or SS and relevant fields
  • At least one UG course in calculus
  • Knowledge in computer programming language preferably object based
  • Bachelors in biological sciences are accepted provided additional courses in CS or SS are taken.
  • Others common for PhD applications in general (LoR, GRE/TOEFL, etc.)

If you go by the above criteria, you ought to get admitted. You may lack other preferences that the institution is looking for. Some may feel that it would take more effort for an non-CS to learn SS than the other way around. Seeing as you stated you do have the skills required, if you are certified in them, you may mail the concerned professors (prospective advisors) of the institution to get a better idea about you taking the program.

  • But I didn't do calculus. I did a bit of calculus in high school though but not in UG. I did go through calculus recently by myself with a text book rented from library but I don't have certificate. Should I redo it from MOOC and get a certificate?
    – Renee
    Sep 2 '16 at 10:00
  • @Renee: I'm afraid so. Nearly all of us did calculus at high school, but this wouldn't have been stated as a requirement in all of the universities if it isn't important.
    – Ébe Isaac
    Sep 2 '16 at 10:12

Although the above comments are proposing to begin with MSc in computer science, I think that it comes back to your motivation and interest for doing the PhD work. As a real story, you can see the story of Carrie Grimes who studied anthropology, then changed to Data science in PhD and finally get a highly ranked job in Google For Today’s Graduate, Just One Word: Statistics , Quora discussion about her.

I knew about the difficult task that you'll confront in your PhD at this major, but you save time and money for yourself.

Before the beginning of your decision, take some online course (or tests) in calculus and specially in statistics to know your abilities. The computational social science deeply depends on network science techniques that is totally correspondent to statistics, probability and computer algorithms knowledge.


Yes, although having a background that involves computing methods (ex. computer science, GIS, statistics, etc.) will make you a stronger candidate.

Computational Social Science is still a very new field. If you look at the CV's of the big names in the discipline very few of them come from a computer science background. It is much more common to see things like geography (Crooks), physics (Grimm), political science (Axelrod, Epstein), and sociology (Carley, Gilbert). Most of the methods in computational social science are not extremely complex to code so computer science skills are not really a barrier to entry.

As always, the best way to resolve the question is to reach out the universities and departments you are looking to study with. Very few universities are offering the degree and George Mason is the only really big name program I can think of off the top of my head. They prefer that student come from either a computer science or a social sciences background which generally seems to be the trend. You may also want to look into various social sciences departments to see if they have anyone interested in computational social science since they might be able to guide you to a PhD in Social Science using computational social science methods.

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.