VIGRE

How Do You Survive Grad School?

A panel of three professors, and three graduate students fielded questions from professors, graduate students, and undergraduates. The panelists were

Mason Porter, Postdoc	Evans Harrell, prof	Michael Lacey, Prof
Sam Greenberg, ACO grad	Todd Moeller, grad	Luis Hernandez, grad

It was a full hour of discussion. Your scribe gives an overview, but does not claim to be accurate to all the opinions voiced. Sam Greenberg stressed that the grad students should talk to peers, and profs about research, advisors, and everything else. This is good advice for those of you reading this.

The Advisor An early topic of conversation was the relationship to the advisor. Whatever an advisor is, the average number of advisors on all those on the panel was, I think, more than 2. While the advisor should be entered into a spreadsheet in the Graduate Advisor's computer. Such entries are easy to change.

So what is the point of the advisor? In Professor Harrell's opinion, one should ``fall in love with the problem, not the advisor." The thesis topic w ould ideally be a rich enough topic to sustain your interest for a period of time beyond your thesis. And the advisor can continue to play a helpful role to the PhD student, with postdoctoral positions, conference invitations, and the tenure track position.

There is no hard and fast rule to choosing the advisor. One indication is recent track record of the advisor in terms of students finishing, and being placed in postdoctoral positions. But, to take a local example, Professor Lubinsky has a very strong record of training students in his native South Africa, but has not yet advised a grad student at GT. Likewise, a young professor, with a hot research record, may not have any record at all, but could be an ideal choice.

Getting Research Done You are in the PhD program to do research. Sam Greenberg pushed the idea of getting into research as possible. Take a homework problem, and twist it, and generalize it, to get a new problem, and possibly a research topic. Attend seminars. Ask questions. Attend the Research Horizons talks, which are targeted at newbies.

Todd Moeller encouraged students to take advantage of range of summer programs for grad students. Todd specifically praised he North Carolina State program in applied math. There is also the IMA program, which was held at Georgia Tech this summer, and the the MSRI has another program. Georgia Tech is a member of the IMA and the MSRI, making participation in these programs especially easy. Participation in these kinds of conferences can help you learn quite a bit about a field, it's questions and future.

[Note: The National Science Foundation's programs are not exclusively targeted at US citizens. Any grad student support that comes through an individual grant, CAREER grants, or Focused Research grants, among many other examples, carry no citizenship restrictions. A prof doesn't even have to be a US citizen to get an individual grant from the NSF. Other programs, such as the VIGRE and REU programs do carry citizenship restrictions. These programs were formulated in part to respond to national interests, and congressional will. ]

One stumbling block to getting research done is the difficulty in reading a research paper, say one that your research advisor gave you. Some strategies are:

• The paper might be cutting edge material, that is written at a very specialized audience. You may not yet be part of that audience.
• Flip to the back, and find the key original references. These are more likely to explain the context and importance of the problem at hand. You can also look for recent survey articles, and textbooks and research mongraphs that address that topic.
• Find out what the key motivating examples are. And study those. The advisor probably knows what these examples are.

The fact is that reading research papers is hard. You can also try to identify the unorigonal bits, the parts that are standard aspects of the subject, and the orginal arguments that represent the contribution to knowledge.

But what about the Comps? Sure, you can't just leap into research. You do need to learn more--Real Analysis, Algebra, and the like. And you need to pass the Comps.

This is a conflicting goal with doing research. There isn't a good answer to this one. Not everyone should pass the Comps, since to do well on a thesis, you will need to be a good problem solver, among other things of course. And we all have a range of conflicting demands on our time. Try balance them out as best you can.

There is a specific timetable for passing the Comps. At the same time, there is no expectation that one rule will be fair to all cases.

Posted by lacey at August 26, 2003 04:55 PM