This blog is about all the important lessons that I learnt while applying to the graduate school. I wish to discuss certain ideas to be kept in mind when applying for Masters as well as Doctoral studies. I hope it will help juniors clarify some of their doubts when they sit down and apply to graduate school.
1) The importance of a good GPA – Almost all the top notch graduate schools look for students with a very high undergraduate GPA. It is nearly a very essential pre-requisite for getting admitted to a good school. A GPA ~ 9.5 and above is considered to be a good one (in my opinion). The admission committee looks for students that are willing to work hard and are disciplined in their approach. GPA is a good measure. Think of a dozen professors trying to figure out the skill sets of a thousand applicants. The first obvious filter has to be GPA.
My GPA around the time I had applied for graduate school was 9.00. My departmental rank around 8-10/60.
2) A good research profile – As far as admission to masters is concerned, a good enough GPA and good research projects might be enough for you to get in. However, if you have a decent GPA (8.5 – 9.5) you might have to supplement it with some good research projects. Publications are standardized metrics on which measurement of a student’s research interest is generally done. Any student with a decent research experience (a couple of publications) would actually hold him in a very good position. Students with high GPAs might lose out. The reason is pretty simple. Professors are happy to bring in research talent, who could actually go on and become research students under them.
As far as doctoral programmes are concerned a few publications are almost essential (unless you are amongst the top students in your batch). Be it student symposiums, posters, or short papers; the research work might be in a second or a third tier conference as long as you have put in a decent research effort things might just work out. Try to be first or the second author. Otherwise, the impression is that you might have just helped out in the experiments and there was no real significant contribution from your side.
By the time I had applied for graduate school I had a decent research profile. This was the biggest change in my graduate school application fromt the previous year. I had 3 IEEE publications, 1 poster (LNCS), 1 student symposium presentation, 1 crappy journal publication, 1 LNCS paper (all first authors). This made up for my below par GPA. The admission committee looks at what level of dedication does a student possess. Therefore, it becomes imperative to complete projects and get them published in some conference. Conferences tend to bring in experience enthusiasm as well as self-confidence. A good publication record more or less makes up and can really help you out with your applications.
3) Recommendations – I believe this is one of the most essential aspects of your graduate school application after your own personal profile. It is essential to build a good relationship with your professors. The best way is to work on some research projects under them.
My recommendations were from a professor under whom I had interned, my BTP guide (had 1 publication with him and 3 courses), my other guide (under whom I had 4 different publications), and a professor (under whom I had two courses). All these professors knew me decently well and I believe gave me positive recommendations. The admission committee requires good recommendations rather than decent ones from well know professors. Thus, it is imperative that you prove it to your professors that you can do decent amount of research and have genuine interest in it. Professors are very helpful and encourage students to study ahead. A reply from my professor read something like this :
“Nothing brings in more joy for a budding teacher like me than my students getting motivated to study further.“
4) Choice of graduate school: While most of the students apply to all the top notch universities, it is essential for one to go through various research labs of different institutes who are doing cutting edge research in the area of one’s interest. This could take anywhere from 1 week – 3 weeks to have a finalized list of graduate schools. For a PhD it becomes very essential to look at the profiles of professors before applying to them. A very high match of interests can really help you get an admit under him, which makes a lot of sense too. A student interested in HCI should go for CMU, MIT media labs, etc.; for theory Princeton; for systems UC Berkeley etc. It is essential to do a through analysis of schools, labs, graduate students to the extent of analyzing your own chances.
Schools tend to pick students from the same university year after year. Thus, getting in touch with your seniors help; even to the extent of nagging them might work.
I started the ground work by July and by the end of August I had a decent list of schools. I almost applied to all the graduate schools where my batch mates had gone into the previous year. This meant that my chances were not too bleak as I already knew where I stood in comparison to the other students. I applied to 13 different schools (for 14 courses).
MIT (pending), UCB (pending), CMU (reject), Princeton(accept), Cornell (pending)
UT-Austin (pending), UIUC (pending), UW-Madison (pending), University of Southern California (pending), University of California Los Angeles (pending), Purdue University (accept), University of Massachusetts Amherst (pending), Stanford (pending), CMU (reject), Georgia Tech (pending).
I had applied for specialization in Computer Systems (esp. wireless sensor networks, except for Madison where I had applied for File Systems).
5) Statement of Purpose: I think another important change in my graduate application from the previous year’s application was the fact that I was pretty clear with my intention of pursuing my graduate school application. Once you are clear it becomes easy to write what you wish to do. I also did some basic research on what to write in a statement of purpose (Link  to some filtered advice). The underlying idea was to convince the admission committee that you are genuinely interested in research (in my case). A good statement of purpose might take anywhere between 1 week to 4 weeks, to refine it and get it reviewed by others.
Link  to my SOP. I took almost 2 months (mostly working on weekends) to prepare my SOP. I also got it reviewed with some of my friends.
6) A good online profile: I believe in showing what you have done. A decent homepage will do it for you. This way, the committee could actually cross check all the information that you have put in your application. A lot of information may not be possible to be written in the application. Your homepage could be an infinite space where you could show case your talent. Have profiles on sites such as academia.edu too. A lot of colleges specifically ask you for your homepage links. Thus, it would really help if you could portray your work better.
7) Good research projects: I did all my internships outside IIT-Guwahati, which I came to regret later on. My internships were @ Adobe India (2nd year) and University of New Hampshire (3rd year). They helped me to experience good quality work. However, the biggest drawback of such 2.5-3 months internships is that it is hard to get any decent research work out of such internships. My advice would be to stay back in summers, work with professors there @ IIT and stick to those projects till you get something published. It is not that hard after all. That way you could build a good relationship with your professor. You also have a great chance to work independently on a project you choose. Undergraduates are never expected to publish papers in top tier conferences; decent research projects with credible results and some publication will suffice.
All my publications were projects I did in IIT G itself. Most of my major work was done after I got all my rejects for 2012 fall semester. I had to stay back at IIT after the college ended , I also stayed back some winter vacations partially before too and by the end of June 2012 I had finished my projects. My first project had taken almost 4 semesters to complete. Hence, it is essential to stick along and have faith in one self. We are good enough to get through the struggle at the undergraduate level.
8) GRE – TOEFL: Graduate Record Examination (GRE) and TOEFL are basically test of very general aptitude of a student in English. The amount of time required for GRE could be anywhere from 1-3 months (along with college studies) and for TOEFL 1 – 3 weeks. A decent score is good enough to get into any good college. Some college require a high score. The basic purpose of these tests is to test your english speaking and comprehension proficiency. In the graduate school one has to go through innumerable research works which pretty hard to understand. Besides, good writing skills make it easier to put your ideas in writing and get them published. Hence, it becomes essential that you prove your english proficiency. Though a bad score should not hamper your application too much. So, I guess one should not get very paranoid about it.
GRE -Subject – I did not give one. You could take one, it would supplement your application. However, it is not necessary.
I spent a month preparing for GRE (my score was 158 Verbal, 163 Quant, 4.0 in analytic writing. TOEFL – 112/120).
9) Building a research profile: Working on a research area right from the start can be very helpful. If you could narrow down your interest into some specific domain and pick on a problem, it could expedite your progress within that field. One could start from the topmost level say Computer Systems, Machine Learning, Theoretical Computer Science, etc, then zoom into a research area. This would help you build deep insight and understanding into some specific problem. Your academic profile would reflect genuine interest. A diffused set of projects in various fields might not be very useful though.
I started working on Sensor Network Simulations during my fifth semester. Most of my published works are in this field. While applying for graduate school I was able to portray a strong profile in sensor networks. I had 3 publications in sensor networks. This played a vital role.
10) Miscellaneous: I do not find contacting professors over emails useful. Though if you already know a professor already, it could help your application.
I tried contacting professors, but did not find success there. I believe you can get through without it too.
P.S – For any queries I am ready to help.
Email – firstname.lastname@example.org
4) My Homepage