Grad School as a Job : Improving your odds

A while back I read an article by a professor about how studying in grad school is like a job. The terms of the job are different from other industry jobs. You are paid to do research. You are paid not just a small salary but also in tuition towards your degree. This is in some sense the perspective of a professor at a university. His top interest is in doing good research in his field, to either obtain tenure, or just to gain respect in his own area of expertise. So the question that professors ask when looking at applicants is "Who will be best able to do my research work?" What is the number one skill in doing good research? Creativity. But a far more basic requirement is having the basic knowledge in the field of research. Communication skills are also very important because finally good research needs to be published. If you write an email to a professor replete with grammatical errors or spelling mistakes, you have almost no chance of hearing back.

So below I outline steps for building up a case to get admitted to a good graduate program with a Research Assistantship. Understand that you don't have to build up case in this manner exactly, if you didn't do things with a focussed goal in mind, you could try to build a case based on what you already have done.

- Look through a department's list to find professors whose research most matches your interests. Hopefully there are other places that do this kind of work, so you will be able to apply to more than a few places.
- Second  you need to a be trained in the basics of the field. For instance say you are interested in Operating Systems, it is important to have read the basic texts in the field. If you have a choice of electives you need to take electives which will get you a better understanding of OSs. You need to have some awareness of the latest research in the field. You'll need to read at least a few papers and understand the content therein.
- Next your task is to choose various projects to develop some expertise in Operating Systems. Your final year project and third year project are 2 examples. You could also get involved in part-time projects like fixing some bugs in an Open Source project like Linux.
- You could next present papers on related issues at student paper contests. These probably don't have as much impact as publishing in a journal or a well-known conference. If you are able to pull that off (and I know that's very hard with the facilities at some colleges), you are virtually assured a spot in a good program. Publishing a paper in a reputed journal is the actual job you are being hired for, and you have already distinguished yourself from most other applicants in that regard.
- Another major criteria that admissions committees are trying to judge is intelligence. GRE scores, GPAs are all essentially a way to measure intelligence. The better your academic performance the better your odds.
- Selectivity or reputation of your undergraduate college is a big factor in the decision making process at top-universities. I realize that unfortunately there are only a few famous universities that an average professor in the US will have heard of. In India these are usually the IITs, so most of us non-IITians are disadvantaged in this regard.
- Industry experience with a reputed multinational in a closely related field should also boost your chances of  entry. The factor that could give you a big edge over your peers here are patents, which are universally accepted as a sign of innovation.
- Working with a famous person in the field also is another way of getting into a top program. I knew someone who did a short research stint with a professor at IIT Bombay in Databases and later went to a top 10 CS program based on his recommendation.