Sunday, March 16, 2014

Grad School Decisions

In the computer science PhD world, prospective students are just about wrapping up their school visits, and they will soon make a decision on where to spend the next 5-6 years of their lives. The decision can often be a difficult one, and typically must be made with imperfect information, conflicting advice from others, and uncertainty about how to weigh the various influencing factors.

In my opinion, the **single biggest indicator** of grad school success is how well you work/fit with your advisor. Ben Barres recently wrote a great article expounding on this very issue. I won't bother with repeating all the great points he made, but I want to emphasize some things from a computer science perspective.

  • It's important to work with an advisor who has an exciting vision, rather than just doggedly work someone whose research area matches the narrow scope of what you worked on as an undergrad. Keep in mind that a significant fraction of students (perhaps as many as 50%) switch areas after starting grad school from what they declared in their applications. Of course, most switches are to neighboring areas that have significant overlap in the technical foundations. But the lesson to take away here is that you should keep an open mind about what research you might find interesting.

  • It's important to have a good working relationship with your advisor. Research is a very unpredictable process with a lot of highs and lows. You will almost invariably hit some bumps while working with your advisor, and it's important to have a good rapport with your advisor when working through those bumps. Assessing this somewhat intangible "fit" can be hard to do a priori, but simply talking to a potential advisor can often yield some warning signs or illuminate that there's a good match. Talking to other students who currently work or have previously worked with the potential advisor can also be useful. But beware of students who have no first hand experience with the potential advisor -- sometimes students just like to say things =)

  • Some professors are more hands on than others. Some professors like their students to work in groups, while others like their students to work alone. Some professors want their students to take a deep dive for 1-2 years on a grand project, while others want their students to be more focused on the next 4-6 month project. All of these work styles can lead to successful research outcomes, but it has to work for both the advisor and the student. And, of course, some professors are more flexible to adapt their advising style to match the student than others.

  • Working with the right advisor is usually more important than quibbles over funding. Most schools guarantee funding for PhD students. But if your advisor is low on funds, then you may have to spend a few extra terms as a teaching assistant rather than be funded directly by your advisor as a research assistant. In my opinion, this is completely worth it if it's for the right advisor (the one caveat being that the school should have reasonable TA workloads). However, if your advisor is completely broke and can't fund anything (including machines and travel), then that can become problematic.

  • Good luck to everyone making decisions!

    3 comments:

    MD SHAHA JALAL said...

    Most of the accounting packages have customer statement version available out of the box.  There are reasons for you to decide recreate the statement in flexible design tool. See more amcas personal statement

    aliya seen said...
    This comment has been removed by the author.
    aliya seen said...

    Really good post! I would like to say about it that it is very informative post. I really like this post. Keep updating me academic project writing