akrish

Sometime in February or March you’ll start hearing back from schools. Unless you are incredibly brilliant, you’ll probably get into some and not get into others. If you only get into one school, and you’re set on going to grad school, then you don’t really have much of a decision to make. I guess you could decide whether that school is good for you and if it isn’t look for a job or something. This article won’t talk much about this scenario because I don’t know anything about it.

If you got into more than one school, you have to decide amongst them. In computer science, most schools will invite you to visit for a couple of days (usually a weekend, but not always) to check out the school, the department, and meet with students and professors. I highly recommend going to these visits. First of all, they are really fun, you will meet a lot of new people, include new admits, and you will see the admits over and over again at all of the schools you visit. Secondly, they really help you in choosing schools; if you’re going to spend the next 4-n years of your life somewhere, you don’t want to be making a blind decision. Third, I was mostly done with my undergraduate degree, so academic commitments weren’t really an issue. If you are taking a heavy course load (or otherwise busy) find a way to get your work done so that you can make it to these trips. In other words, don’t use school as an excuse for not going.

I think you should prepare a bit before you visit schools. You’ll probably be meeting with faculty, and you’ll want to get the most out of these meetings to help you make an informed decision. You should be at least familiar with some of their research (read their papers) and prepared to talk to them about it. What you’re really trying to do in these meetings is see if the faculty member would make a good advisor for you. This means you’ll want to be interested in what they’re doing in terms of research, but you’ll also want to connect with them on a personal level. I found that I could connect with professors better if I knew about the work they had done so we had something interesting to talk about. Faculty meetings are probably the most important part of your visits. Really, as a grad student, everything except your research is kind of secondary to your research. And the person guiding your research will be your advisor, so you want to make sure that you choose a good one.

However, other things are also pretty important. I cared a lot about whether I would be happy at a school or not (seems reasonable right?). At some schools I got a lot much better with both the current students and the admits than at other schools, and some schools seemed to be a much better fit for me personality-wise. You should think about these things too, ask students what they do for fun, whether they play sports, etc? If you’re interested in something (like a sport) and want to continue it, see if you can find people interested in the same thing at the schools you’re visiting. These things shouldn’t make or break a school, but if you weigh two or more schools the same in terms of research/advisors, these issues may make a difference.

Another thing you should do is talk to other admitted students in your area. See what they think about the school and if they are considering accepting. This is important because they are all going through the same process and maybe they can contribute some information about something you haven’t realized or thought about. Also see if you get along with those people. This is important because you’ll spend a lot of time interacting with those students and you want to like them.

Some people told me to make lists of the things I found important and be really systematic about my decision making but I decided not to do that. I visited schools and kind of went with my gut to formulate an opinion about them. I can’t really say anything about the systematic approach, but I didn’t really have any issues formulating opinions without it.

Don’t forget that visiting schools should be fun. Most of the time schools plan a lot of fun activities that often involve large amounts of alcohol. Enjoy all of the new people and the new places and keep in mind that you will be fine most places that you go.

next up… deciding!

Here’s what I did with regards to applying for graduate school. Maybe this will only apply for computer science PhD. programs.

First
I got involved in research at the beginning of my sophomore year. That year I didn’t really do much of what I would now consider research, but I showed my advisor that I was a good programmer and a hard worker. Starting in my junior year, he gave me a small research project with the goal of actually publishing the work. We actually submitted a paper and it was accepted to WWW which probably helped me a lot. In addition, I spend the summer after my junior year working full time on a research project.

Doing research asap is important because a. it makes your application look better and b. it helps you decide whether graduate school is right for you. If you’re still young, get involved right away.

The application
Once you’re preparing to apply, it’s kind of too late to get started on a research project. Instead you should focus on your application. Filling out the applications are generally pretty easy; the hard part for you is to write one or more essays. Additionally, schools generally care a lot about your grades and your recommendations.

Academics
Everyone I’ve talked to says grades are important, but your research is more important. Schools generally care to see that you are capable, but it’s not a big deal if you have weaker grades because you’ve been working really hard on research. While you should be getting pretty good grades, it’s more important that you are taking challenging courses and courses that are aligned with your research interests. Taking courses aligned with your interest is important because you will get to know professors in that area, which could lead you to interesting research projects or to good recommendation letters.

Also, if you can, I would recommend taking graduate courses. Generally, you will build a better relationship with your teacher (because classes are smaller) and you will often have a chance to do a research-style project. I asked a teacher from one of my graduate courses for a recommendation because I worked with him pretty closely on my class project. Another graduate course I took pretty much confirmed that I wanted to pursue machine learning.

Jean Yang mentions that if you do have weaker grades you should talk about them in your essays. Saying something about how you spent most of your time on your research would be good justification.

Recommendations
Recommendations are super important. When you think about it, if someone is willing to put his or her reputation on the line to vouch for you, people will take them seriously. Professors generally are candidly honest in their recommendations, because they don’t want to recommend a bad student, which could jeopardize their image. Consequently, if you can get three really good recommendations, then you may be golden.

Most schools ask for 3 letters of recommendation. First of all, recommendation letters should be from faculty members (professors); you want them to have to be accountable. Recommendations from graduate students or from industry (unless from a research lab) don’t really carry much weight. The best recommendations generally come from professors that you have actually done research with. They have the most to say about your research ability because they’ve seen you work. Teachers may not be able to add too much to your application if you haven’t interacted with them much outside of the classroom.

I think it’s pretty tough to get three recommendation letters from people you’ve done research with. Apparently so do people on admissions committees. It’s perfectly reasonable to have 2 recommendations from research advisors and one from a teacher that you had a more-than-normal interaction with. If you take graduate courses you will have a good interaction with your teachers and they will probably be able to write pretty good recommendations.

I got recommendations from my undergraduate research advisor, the professor that I worked with over the summer, and a professor from a graduate course I took. I think this was pretty reasonable.

Essays
Most schools only require that you write a personal statement. This isn’t a personal statement like what you wrote for your undergrad applications. It should really be called a research statement. Applications committees care about what research you did and why that pushes you toward graduate school. Therefore, you should talk about your research experiences and specifically what research you did.

I was talking to a friend of mine who’s thinking about applying in the fall and he was kind of worried that he hadn’t done any cool research. This is pretty normal for undergraduates; I think it’s pretty unlikely that you’ll be doing some cutting edge work. However, there are probably small things that you did that can show off your research abilities. For example, I spent a couple of days research various ways to randomize graphs, just to evaluate the actual algorithm we were working on. I talked about this in my essay and I think it showed that I had some intellectual curiosity.

Your essay should also include some school specific information. Spend some time figuring out why you want to go to each school you’re applying to and write that in your essay. Also look at people who you’d be interested in working with and talk about them briefly in your essay. If anything it shows that you care enough to do some homework on the school.

UC Berkeley also has you write a diversity statement. In this essay, you’re supposed to talk about how your background will contribute to the diversity at Berkeley. I found this essay pretty difficult because I’m an average Indian-American computer scientist. Given that, I talked about the summer I spent in Israel doing research and how I am motivated to collaborate with researchers locally and internationally. Unfortunately, I didn’t get into Berkeley so maybe this wasn’t the best way to go.

The Rest
The rest of the application stuff is straightforward. You have to fill out some personal information, send in transcripts and all of the usual stuff you did as an undergrad. Really at this point most of your work will be in writing essays. However, if you applied for fellowships, you should be able to take a lot of your material from there.

Next up… visiting and making a decision.