Independent Study Guidelines
I love supervising undergraduate students in independent
studies. I've found it useful over the years to develop some
guidelines, and here they are. Please feel free to contact me with any
questions: ferguson@cs.rochester.edu.
Before You Start
So you've got an idea that you think would make a good independent
study. Here are some questions that you should ask yourself:
- Is the idea well-defined? For example,
multi-player
game featuring flying donuts
is not well-defined. Implementing and
evaluating a specific information extraction technique on a specific
set of webpages is well-defined. A good test of this is whether you
can you say what you want to do in a couple of contentful sentences.
- Do you know enough about it? There's no point proposing to
work on something about which you know nothing. Students should
generally have taken the pre-major and relevant core CSC courses
before considering independent study in CSC. If you have an idea but
don't know much about it, do some reading before you talk to me
about it.
- What will you learn by doing this project? In a sense, this
is the most important question of all. Pick something that you
really want to work on, and that will stretch you in an interesting
direction. Unlike regular courses, you get to set the agenda, so use
the freedom intelligently. You should be able to explain to others
what you'll get out of the course.
- Is it feasible as a term project? You only have a few
months in which to go from nothing to something, and you presumably
have other things going on also. Look at what's out there. Make sure
that you budget your time realistically, taking into account the way
you work. A note to seniors and other over-stressed people: Since
there are no exams or assignments, independent studies have a
tendency to be that thing you drop when you run out of time and/or
energy. You don't want that to happen.
- Can you work on it independently? An independent study
means that you study more or less on your own. I can help you refine
your idea initially, and I can help you with suggestions when you're
stuck, but really you need to be able to make progress on your
own. I will not be keeping close tabs on you. At the end of the
term, you need to have something to show for your time.
- Is what you are doing research? Not all independent studies
have to involve novel research, but the most interesting ones do and
I encourage you to find a research angle in what you want to do. In
this context, research means pushing the boundaries of what we know,
trying something new, and evaluating your work either theoretically
or experimentally. Having done something that is publication-quality
is good for your resume and great experience for the next stage of
your career.
- Am I the right person to supervise you? I enjoy
supervising students who have found a research niche close to my own
interests. You can find these from my homepage and my publications
if you don't know them already. I don't mind supervising students
doing other cool stuff provided I know something about it. You
should also keep in mind what I said about working
independently. Other supervisors might be more hands-on.
- Do you want an A? Bottom line: there's no point taking an
independent study and not getting an A. Otherwise someone looking at
your transcript will wonder why you couldn't get an A on a course
that you designed yourself. You can't blame the prof or the exam or
anything else. A solid plan executed diligently will be worth an A
regardless of the final results.
Registration Requirements
In order to sign up for an independent study, we need the following
items:
- A title for your project: While this seems trivial, in fact
it is important. Since it appears on your transcript, it's a chance
to say something that will really catch a recruiter's or admissions
committee's attention. Speaking from personal experience,
Agent
Team Coordination
or Robotic Musical Chairs
really sticks
out amid the sea of Calculus I
and Intro to
Programming
listings.
- A short description of your project: On the order of a
paragraph. Anything longer can also be attached, but you should be
able to be pithy.
- A timeline with milestones: You need to develop a
reasonable schedule for your work. Milestones along the way (at
least monthly) help us make sure things are on track and detect
problems in a timely manner. Milestones can be changed, but by
having a schedule at least we'll be forced to talk about it before
the end of term. Be realistic! Your last milestone should be turning
in a final report for the project on the last day of classes.
I can help you with these, but you should probably try to sketch them
out before coming to me. If you answered the questions above, you should have no problem doing it.
You also need to register for the course and do whatever paperwork is
required by the department. Check with the CSC Undergraduate Coordinator well ahead of time.
Writing Your Final Report
For the report, it doesn't have to be too detailed, but it should read
like a research paper or technical report. You can find samples of
these on my website or on the department website. Be professional.
The overall structure should be something like the following:
- Start by introducing the problem, discuss possible approaches, maybe some previous work if you used it in understanding the problem.
- Then describe what you did, including false starts (which you probably
wouldn't publish in a real research paper ;-)). Things that might
count as technical contributions include background work (such as
corpus gathering), tool or system building (programming), or theoretical analysis.
- Then on to at least some more or less formal
evaluation of your efforts, and comparison with other approaches (in a
real research paper you'd want to show how your method was better than
others).
- Finally describe the things left undone, or that you would do
better or differently (
future work
in a research paper, perhaps
also like lessons learned
in this kind of
report).
- Traditionally one's conclusions repeat the abstract or
introduction (more briefly) in the past tense.
It is very important to include properly formatted references where
applicable.
Using LaTeX and BibTeX makes things look nice and is worth learning
anyway if you don't know it, but it's not a requirement.
This may sound like a lot, but you will have already done some of the
introduction in the proposal. The whole needn't be more than, say, 10
pages, but less than, say, 6 would probably indicate lack of effort.
Deliver a PDF via email. If your project involves developing programs,
make arrangements to give a demonstration and get me the code (both
before the deadlines).
Self-Assessment
At the end of the project, I will ask you to give yourself a grade for
your work. You can provide a short justification as well. I try not to
look at that before reading your report and deciding on a grade
myself. The final decision is mine, of course.
Last change: 21 Feb 2019