Lecture 10: On Being a Graduate Student

• References

• Finding an Advisor

• Reading

• Becoming Part of the Research Community

• Doing Research

• Writing Up Your Results

• Looking for a Job

References

[d] Marie desJardins, How to Be a Good {Grad Student,Advisor}, marie@erg.sri.com, March 199

[S] Survival in the Academy, from Internet newsgroup, source unknown

Finding an Advisor

First identify why you went to graduate school

To teach?

To get a higher paying job?

To postpone your job hunt?

Because you like the academic environment?

To become a researcher?

Learn about the department's activities

• Attend grad seminars of interest - in CS and other departments

• Talk to other students about what they are working on

• Look at the department's technical report series
  (http://ei.cs.vt.edu/NCSTRL.html)

The role of the advisor

Serves as your mentor

Serves as a source of technical assistance

Helps you write the best thesis/dissertation you are capable of, given constraints of time, ability, knowledge, etc.

Helps you find resources you need (financial, equipment, psychological support)

Helps you prepare for oral exam and defense

"Keeps the department and other hostile entities off your back"[C]

Introduces you and promotes your work to important people in your field

Gives you advice on the direction of your thesis/dissertation

Gives you advice on career directions

Helps you find a job

After graduation:

• May provide advice
• Will become a colleague (thus changing the teacher/student relationship)

Adivising styles

Much of this is taken from [C, p. 13]

How much direction do you want?

Some advisors give you a well-defined thesis-sized problem, explain an approach, and tell you to get to work on it.

Other advisors are hands-off: they may give you no help in choosing a topic at all, but can be extremely useful to bound ideas off of once you find an idea.

So do you work better with a lot or a little structure?

How much contact do you want?

Some advisors meet weekly, possibly in groups, for progress reports. Others meet on a demand driven basis. Some might want to meet once a semester.

How much pressure do you want?

How much emotional support do you want?

Who will act as a filter for ideas?

Some advisors toss out lots of ideas and expect you to figure out which ones are good ideas. Other advisors will try not to overload you with ideas.

Is the advisor willing to supervise a thesis on a topic outside his/her main research area?

Finding the right advisor

• Introduce yourself to each faculty member whose research areas fit your interests

• Visit research different group meetings

• Ask your peers about their relationship with their advisor

• Find out how many students a faculty member advises

• Find out the average, variance in time to complete a degree

• Find out how often students drop out

• Look at the theses/dissertations of past students

• Find out where the advisor's students have gotten jobs

• "The advisor-advisee relationship is necessarily personal, and your personality quirks and your advisor's must fit well enough that you can get work done together." [C, p. 13]

• Even "interview" each potential advisor before making a choice!

Changing advisors

Sometimes things don't work out.

Sometimes an advisor leaves the university.

Sometimes you hate an area and discover you love another.

You can change advisors.

• A reasonable time to change is between the M.S. and Ph.D. degrees.

• Weigh the time required to start over with a new advisor agains the risk of not completing your degree with your current advisor

Reading

Most of the initial work on a thesis/dissertation is reading.

You'll be reading textbooks, journal articles and conference papers.

Start by reading the "classic" papers in your field.

A Ph.D. student will read paper drafts in later years.

You may find published bibliographies for your area (e.g., computational geomtry)

I also suggest browsing the library for a couple of hours late on a Friday afternoon just to let new ideas pop into your head.

When reading a paper (from [d, sction 3.1]:

Make sure it is worth reading before investing the time

First skim the paper to get an idea of what it is about, without investing days in understanding notation or filling in background material

If it is still worth reading, reread it carefullly.

Make a notation list while reading.

Some questions to ask:

Has the idea been tried on real data? Toy examples? No examples?

Are you reading buzzwords and jargon or deep ideas?

Is this paper a small increment in knowledge or a new idea?

Becoming Part of the Research Community

You should actively interact with people in your field outside of the university. To do this:
• attend conferences - introduce yourself and/or present a paper

• send manuscript drafts for comments to people you meet

• offer to comment on manuscripts of other researchers

The Daily Grind

What do you do each day as you are actually "doing" your thesis/dissertation?
Reading papers

Discussing ideas with your advisor andother students

Staring into blank space trying to devise a solution to a problem

Staying Motivated

It is easy to loose months of time with no real progress to show.

It is also easy to get tired of working on your thesis/dissertation problem.

You may also read a dead end or a point of frustration.

Some ideas (from [d, section 3.2]:

Talk to someone when you're having trouble (advisor, friend)

Set up regular meeting with your advisor

Through out an unrealistic schedule ("Finish 1 chapter per month")

Find the fastest method to test out an idea or theory, without investing months of time to find out it is wrong.

Focus on achieving a well defined, small goal (possibly leading to a chapter)

Notebooks

The motto for some fields of science: "If you do not write it down, it did not happen."

I personally have a set of notebooks going back 13 years with all work I've done on the main thread of my research.

Notebooks are essential for experimental work -- it is so easy to get confused as to what you observed!

For CS research, the notebook can be inches of computer printouts with trace data from experiments.

In fact, I even keep trace data output from student theses and journal papers in case I need to see why we drew a conclusion years later.

Also put your ideas in notebooks as they come up.

You should try to write as you do your research -- rather than sitting down one day with all the results and starting to write.

You might try giving a talk about your work before writing it up.

How to avoid writing a thesis/dissertation:

• Write a technical report

• Turn the technical report into a conference paper

• Turn the conference paper into a chapter, or turn the conference paper into a journal paper and then into a chapter

• Add an introduction and conclusion to the papers you wrote, and staple the whole thing together.

Some hints on writing...

When you "see the light" in how to solve a problem:
Write a one page table of contents for a strong paper that you would like to write

Then do exactly the research needed to fill in the holes in the paper

Then write a detailed outline. In fact, I suggest making slides for a talk and giving the talk to whoever will listen

Refine the slides.

Further refine the slides to make a detailed outline (e.g., one bullet per paper paragraph)

Then write the text. I find that you wind up deviating from the outline in certain local parts

Then put it away and reread it, or give it to colleages to read

Your advisor should help, especially for academic jobs

There are internet mail lists and newsgroups

Attending conferences helps makes contacts

Contacting friends at companies or other former students of your advisor helps too

Sending a resume is the method of last resort