HTML generated by txt2html via TOM

EMIGRATION COURSE: Tips on the Interview Process

Jeannette M. Wing

2/9/96, 2/7/97, 1/30/98, 12/4/98, 2/18/00

Why I Am Giving This Talk

To help you.
To help CMU SCS.
You represent not just you, but also your advisor, SCS, and CMU. * You are our ambassadors.

Outline

Pre-Interview
The Interview
General Dos and Don'ts
Typical Structure
The Job Talk
1-on-1
Department Head
General Things to Keep in Mind * Post-Interview

Assumptions

You've sent your packet out.
You've gotten a phone call inviting you to visit.
You're sincerely interested in the place.
Don't waste your time if you're not. * Don't waste their time if you're not.

Pre-Interview: Dos and Don'ts

Do your homework.
Practice your job talk (at least) twice.
Make up a list of questions to ask (more later...).
Be prepared
To explain your work to different audiences (more later...).
To answer some tough questions from them (more later...).
Bring a notebook (paper or electronic).
Pack some presentable clothes and shoes you feel comfortable walkingand talking in. Think about the location of the place.
Gentlemen: bring a coat and tie. * Ladies: bring a dress or skirt.

Do Your Homework Goal: Know your audience. Find out

Strengths and weaknesses of the place
Who's on the faculty/research staff, especially the bigwigs, majorplayers, eccentrics, ``difficult'' ones
How big (number of people) the place is
A bit about the place's organization
School: private vs. public, dept/school/univ relationship
Lab: How broad and deep is the hierarchy? Matrix?

From

Your advisor
Other faculty or people who went to school there, taught or worked therebefore coming to CMU
Fellow students who have visited there
Friends of above
WWW * Glossy brochures

Practice Your Job Talk

Practice, but don't overpractice.

First time: It will be the worst talk you ever gave and ever willgive in your entire life. (Get some friends to play the

role of known ``difficult'' people.) Second time: It will be pretty good. Third time: It will be great. Nth time: It will bore you and the audience.

Number your slides.
Bring blank slides and pens. * The talk itself (more later...)

The Interview: General Dos and Don'ts

Be yourself
Mind your manners.
Be polite.
Be respectful.
Don't offend.
Don't be (unnecessarily) argumentative. If you disagree with someone, turn it into a fun technical debate, not a religiousargument.

More General Dos and Don'ts

Show conviction, passion about something.
Have an inner voice, a rudder that steers you. Know yourself.
Show an interest in what people are saying, but
Don't try too hard to please.
Don't be too agreeable. Don't be spineless. Stand up for what you

believe in.

Listen carefully to what people are asking or saying before answeringquestions or responding to comments.
Don't say anything stupid.
If you don't know anything about something don't pretend that you

do.

Don't be glib, especially with people you don't know.
Keep detailed notes (peoples' names, impressions, etc.). * Dress neatly.

Structure A typical two-day interview:

Arrive the night before. Dinner maybe.
Day 1
Breakfast.
1-on-1s (30-60 minutes each).
Talk.
Lunch.
1-on-1s.
Dinner.
Collapse.
Day 2: Repeat Day 1 minus talk, maybe minus dinner.

Somewhere in there:

Tour of department (offices, labs, etc.).
Talk with department head.
15-30 minutes private time to gather thoughts, go to bathroom. * Jot down notes (before you go to bed).

The Job Talk Two main purposes

To sell yourself (sales talk).
To sell your research (technical talk).

There are different audiences in the same room

1-2 experts, people who know what you're talking about.
Everyone else.
Faculty/researchers in a tangentially-related field. (How can your

work help me?)

Faculty/researchers outside of your field. (Do you sound like you

know what you're talking about? Does your research problem soundinteresting, worth solving?)

Faculty/researchers who are known to be ``difficult.''
Graduate students. (Watch out for some of them!)
People out of touch with research, e.g., (some) administrators, old

fogies.

I assume you know some general-rules-of-thumb about

Giving a talk.
Making slides. ...so what follows are just some reminders...

The Talk Itself Memorize

The first couple of sentences
Transition sentences (between slides)
``Dense'' sentences (e.g., with tricky definitions of highly technical terms)
Catchy phrases

Before the talk

Look at the audience.
Take a deep breath.
Relax. (Smile!)

During the talk

Use eye contact.
Pay attention to pace
Pace yourself.
You set the pace. It's your talk. You are in control.
Use pauses effectively.
Use feedback: head-nodders, puzzled expressions, blank stares.

At the end of the talk

Say ``Thank you'' (or something that indicates you're done).

After the talk (or later that night)

Reorder your slides.
* Clean them, if necessary.

Bring slides as a backup for any laptop presentation you plan.

Dividing Up Your 60-Minute Slot

First 15 minutes
Motivate the problem you were trying to solve.
Clearly and succinctly (for a general audience) state the problem and

explain your solution or approach.

You need to convince everyone that what you did is interesting and

worth their time listening to.

Next 2 minutes
Start diving into technical talk.
Next 20 minutes
The technical talk.
Go into enough detail so that the experts can follow everything you

say and are absolutely convinced that you've done good, solid work.

You may lose the rest of your audience, but they should still be able

to make sense of your high-level arguments. Also you want toimpress them with something that they don't understand.

Next 3 minutes
Begin wrapping up.
Next 5-7 minutes
Related work: Be scholarly. Explain how your work relates. What's

new? What's different? Don't just give laundry lists.

Last 3-5 minutes
Conclusions, future work (if time). Leave time for 5-10 minutes worth of questions interspersed or at end. Have backup slides for anticipated questions, further details about tricky or interesting technical points.

Multiple Talks

Have multiple versions of the talk in your head, for different audiencesand different durations.
The one-hour job talk version.
The technical one-on-one version.
The non-technical one-on-one version.
The dean/department head version.
The elevator version(s)
The ``waiting for the Wean Hall elevator'' version.
The ride up/down the elevator version. * Refreshen your job talk from time to time. It's going to get stale.

1-on-1: Questions You Might Be Asked Easy

Why did you do what you did for your thesis research?
Why did you use your approach and not something else or someoneelse's? Know the assumptions and limitations of your approach and

solution.

What's so interesting about your thesis research? What's novel aboutyour contribution? Why should I be interested in the problem or

solution?

What's the key insight to your solution? Your secret weapon?
What're the one or two most significant contributions you feel you havemade to the field, to Computer Science?
What difference is your solution or approach going to make to someonewho is outside of your field?

Tougher

What do you want to do next?
What do you see yourself doing in three, five years?
Where do you see the field being in five years? Ten? What are yougoing to do that will help us get there?
What do you think are the top two or three problems in ComputerScience? What are you going to do that will help us solve them?
What do you think is the most significant advance in Computer Science(or in your field) in the past year? Past two-three years?
Do you know anything about X? (Be careful!) * What do you think of X? (Be careful!)

1-on-1: Questions to Ask Definitely

What research are you doing? (Get a feel for how ambitious a researchproject can be at this place; the scope of research activity at this place;

whether there's anything going on of interest to you; potentialcollaborator?.)

Do you have any students? What are they doing?
Do you collaborate with anyone? What are you doing together? Iscollaboration encouraged?
How is your research funded?
What courses are you teaching?
What are the (undergraduate, Master's, Ph.D.) students like?
What do you perceive the strengths and weaknesses of this place tobe?
Where do you see the department heading?
Do you like it here?

Maybe * Are you happy?

With the Department Head (or Dean) Statistics, financial matters, and procedures.

Find out about
Facts about the place
Future plans, short- and long-term
Evaluation and promotion processes
How faculty are funded for research
Any special research support for junior faculty
Role of junior faculty in getting research money (NSF, ARPA)
View on collaboration
Teaching load
Computing facilities support: who buys, who maintains.
Benefits (health, dental, retirement, tuition exemption, etc.)
With respect to you:
How are hiring decisions made? (So you know when to expect to hear from someone.)

Ask Host or Anyone When Appropriate Find out about

Getting students, quality of students, support for students.
Educational programs at all degree levels (B.S., Master's, and Ph.D.). * Expectations of junior faculty or junior member of research staff.

Meals

If you have any dietary restrictions, speak up.
Mind your table manners.
Relax, be yourself, but don't get drunk.
Be prepared to talk shop. Some faculty/researchers will be able to talkto you only during a meal; they might miss your job talk. (And some just

might like to grill you to wear you out!)

It's a good time to bring up social issues, e.g.,
life on campus, life in town/city, housing, schools for kids, two-body

situation, outside interests

but don't ask about these too much unless
you get the sense that they really want you, or
you can't live without something or without being able to do X. * It's a good time to hear the real ``scoop'' on a place.

Time Leftover?

Ask to talk to some graduate students. * Get an informal tour of campus, neighborhoods where you might live,town or city, (Show an interest in your surroundings.)

What They Are Looking For

Theory vs. Practice
Are you a theoretician, a systems person, a bridge person?
Do you prove theorems or build systems?
Are you an applications builder?
Creativity
Are you full of ideas, an innovator, a visionary?
Are you a thinker?
Are you an incrementalist, an integrator?
Are you a clone of your advisor?
Independence
Are you an independent thinker?
How well do you work on your own and with others?
Brainpower
Are you smart?
Are you a clever problem solver?
Technical skills and ability
Are you an engineer, a technician?
Are you a detail person, a techie, a hacker?
Can you program? Can you do math?
Articulate
How well do you express yourself orally and in writing?
Teaching
Can you teach? * What can you teach?

What They Are Looking For, Cont'd

Fitting in
Do you complement interests of current faculty/research staff?
Do you fill in a hole or overlap?
Does your personality mesh well with the place?
Would you make a good colleague?
Are you a superstar?
NSF ``PYI'' material?
ACM Dissertation Award material? * Future Turing award winner?

What You Should Look For

Research
Is there any interesting research going on there? Can you imagine

doing the kind of research you'd like to do there?

Is there someone with whom you can have a deep technical talk?
Colleagues
Number and quality: How many are there? Are any of them any

good? How good is their best? Are you smarter than all of them?

Do you like the people?
Is there a potential buddy?
Students (at all levels)
Number and quality: How many are there? Are any of them any

good? How good is their best? Are you smarter than all of them?

Where are they from? Mostly foreign or domestic? Where did their

graduate students get their undergraduate degrees from?

Teaching
Are there courses you would enjoy teaching?
Is there flexibility in choosing what to teach?
Is the teaching load acceptable? Flexible? Can you ``buy out'' of

teaching?

Management
Do you get along with the department head?
Can you imagine working within the department/research lab's

organizational structure?

Location
Can you imagine yourself living near there?
Think of your day-to-day life, more than where you'd like to spend your vacations.

Post-Interview

Keep in touch with your host or department head.
The chance to write a formal ``thank you'' is in the cover letter whensending back your receipts for a reimbursement check.
Use peoples' names. * Don't pester people about status, but don't let too much time go by.(Show that you're still interested, a ``live'' candidate.)

Congratulations! You got an offer...

You may ask for a second visit, maybe to bring an SO/spouse to visit the place, city, etc.
Don't be ridiculous in your requests, especially what you want put inyour offer letter. It comes off sounding petty.
Do get in writing
starting salary
starting date (this is when your tenure clock starts and sometimes

affects when benefits kick in)

support for computing facilities (for office, home, traveling; to start up

a lab)

support for summer(s) (how many months for how many years)
support for students
support for moving expenses
release from any teaching responsibilities (how many semesters)
any special deals, e.g., using your n years of post-doc, industrial

experience, etc. towards your tenure clock.

Ask about, and maybe get in writing if you sense you need to
secretarial support
policy about if you haven't finished your thesis after you start
Don't ask (now) about support for telephones, the size, location, or paintcolor of your office, an office with a window, a room with a view, office

supplies, parking, etc. You will sound silly. You may ask later, when itis more appropriate. * Negotiate as high a starting salary as possible. Subsequent raises arepercentage increases.

26 Academia vs. Industry

27 Remember Whether you get an offer or not,

You have made new, important contacts for the future.
You have represented CMU and your advisor. Whatever impressionyou gave them is a direct reflection of CMU's Ph.D. program. (Don't embarrass us!)

Finally

Be yourself.
Mind your manners.
Enjoy it.