;; $Id: grad-fellowships.txt 2359 2004-04-23 02:09:34Z quarl $

GRADUATE FELLOWSHIPS

Notes by Karl Chen <quarl@nospam.quarl.org> 2004-04-22 17:00 -0700 (PDT) (Thursday)

Speakers:

Sheila Humphreys (EECS Center for Undergraduate Maters)
Allen Tsao (NSF winner 2003)
John Tsao (NSF winner 2004)
Katherine Yiu (NSF winner 2004)

Generally speaking, all EECS grad students are supported (500 grad students at Berkeley). Exceptions are people that have been here 12 years and the faculty are nudging them out. Some departments are better than others (CS Theory doesn't have the kind of deliverables that industry wants so is not as well funded).

Resources

Berkeley grad: http://www.grad.berkeley.edu/fellowships/
Berkeley EECS grad: http://www.eecs.berkeley.edu/Programs/grad/fellowships.html

Fellowship kinds

National extramural fellowships - you need to pro-actively apply for them
School fellowships for specific research area (e.g. MEMS) are assigned by professors to top N students per area

Extramural Fellowships

Best 3 to apply for:

NSF - 3 years
NDSEG - 3 years
Homeland Security Fellowship - newly created

Hertz

only give out 15 per year (including physical sciences, engineering)

5 years

interview

AT&T

others are targeted:

Packard for black people
others for women/minorities

Why get an (extramural) fellowship

Fellowships are portable - can take to any school
can defer 1 year or take a year off in the middle
higher stipend - $27k per year for NFS; higher for NDSEG
1 conference with travel

Fellowships from companies:

they might offer you positions, but won't force you

Requirements

transcript
personal statement
statement of research purpose
letters of recommendation (4)
GRE scores

Essays

4 essays:
- what's your research plan (hardest)
  - the essay prompt sounds like you need a specific plan, but not like that
  - hard for an undergrad senior
  - you can write about a creative idea based on undergraduate research
  - be interested in research, have interesting ideas
  - if you've done research, start there and expand it futuristically
- personal statement (what will you be doing in graduate school)
- how will your work impact society
JohnT's essays and NSF's evaluation of them are in Susanne Kauer's office

Advice for getting a fellowship:

web page tells you what you need, how to write essay
keep on top of deadlines (have someone make you) - don't leave it until middle of October; start the process early
- due earlier than schools
- schools usually Dec-Jan
- fellowships due first week of November
- 4 letters of rec. instead of 3 for schools - so need to find profs earlier - but they're not due until Dec.
- 4 essays
based on merit, not really financial need

Essay:

demonstrate clarity of thought - not that you wrote the essay the night before
- demonstrate you know how to explain something, how to write
- have people (advisors, peers) proofread it
NSF uses same (or very similar) essay topics every year so you could start that now

Letters of recommendation:

Have a couple of strong ones - not end of world if not all four are good research-professor letters
How to help faculty write a good letter for you?
- give them statement of purpose, transcript of classes taken, up-to-date résumé
  
  Think of this as your portfolio for applying for an award, etc. Get your documents in order!
Berkeley Career Center Letter Service is OK; helps professors since they are busy
When to ask?
- give them some warning ahead of time; then ask in fall
- give them a month
Update if things change (you got some award, VP of HKN, etc.) and update your letter writers

Transcripts

There's a long wait when it's close to deadline, so you have to wait in line for expensive ($15 instead of $4) transcripts. So better to do it in the summer since it takes 1 month.

Intramural fellowships

some professors want to wait for students to pass prelims to know how good they are
other professors scout around to build up their group

Other advice

TAing is OK - helps you learn material yourself
There's a lot of noise in admissions, so don't be too disappointed if you don't get into a certain school (a lot of Berkeley faculty weren't admitted to Berkeley)