Advice for Graduate Student Advisees

How to deal with me: Contents:

Qualifying Exam

The passing the qualifier exam only depends on the score. Determination of who passes the qualifier is determined only by those responsible for preparing the exam. Evaluating research Potential is now a separate step. There is now a separate determination of having demonstrated research potential.

Research Potential

Evaluating research potential is now separated from the Qualifying Exam and consists of two parts:

Candidacy Exam (General Exam)

The form for the Candidacy Exam is as follows: Fill out the graduate school form. Check with them on when it needs to be filled out and filed with them.

Give me a list of courese you've taken in graphics (and related areas). Annotate any 788s, 793s, etc., listing as much as possible the topics and projects you were involved in.

Typcially (but it varies from professor to professor), major exam is week long takehome and minor exam is long weekend takehome. My questions will typically include a question on advanced rendering, a question in your area of interest for your research, and a question that is some type of open question in computer graphics. I expect your answers to be typewritten (word processor). You can search the literature for material unless explicitly instructed not to. You shouldn't really have to concern yourself with length, but for very general length guidelines, I would expect at least a couple of pages for each answer and probably no more than ten pages.

I don't expect you to discuss the questions or answers with anyone else. After the exam, I don't expect you to discuss the details of any answers with anyone who has yet to take the test.

Ask all members if they want copies of all the writtens. I want a copy of all of them and the grad school rep gets copies of all the writtens.

For the oral, prepare a 5 minute 'intellectual history' of yourself in order to introduce yourself and calm down. The biggest mistake you can make is to get rattled. If you don't know the answer to something, admit it and go on - it won't flunk you (usually).
There is a separate presentation of a research proposal required by the Department - see Research Proposal below
It's your responsibility to schedule a time that is acceptable to all committee members (except the grad school rep) and to reserve a room (usually one of the CSE conference rooms, but DL680 is also possible). Some students bring donuts or other food to the general exam. I don't encourage this; you have enough to deal with already. I do recommend having coffee at the exam.

Master's Thesis

The thesis option for the Master's degree is a reasonable way to go. Sometimes students opt for the thesis because they want to leave here with some kind of significant project they can show that they have done. Often, the project can just be an independent study project, but some students want something more significant than what independent study usually produces. Graphics lends itself to theses projects. In general, you should expect a thesis to extend your stay here by one or two quarters. On the upside, it replaces the Comprehensive Exam.

To qualify as a Master's Thesis, the work should be something new, but not necessarily a quantum leap in the state-of-the-art. Being 'new' can take the form of a new survey or taxonomy of existing material, or it may be some piece that contributes to a Ph.D. dissertation. Personally, I tend to discourage programming projects that don't really contribute to either the computing environment around here or contribute to some larger research project. I discourage 'throw-away' programming projects which don't really contribute anything useful. Theses require a lot of resources (both of the computing environment and of myself) so they should be worth the effort.

Graduate Studies Committee: A student who wishes to exercise the thesis option for the MS degree must satisfy the following conditions: The student may retake one or more core courses in order to satisfy these conditions; any previous grade will then be ignored.

A form listing core course grades and the proposed faculty thesis advisor must be signed by the advisor and by the chairperson of the Graduate Studies Committee no later than the quarter preceding the quarter in which the degree is awarded.

Selecting a Thesis or Disseration Topic

Unfortunately, I don't have any pat answer for 'what makes a dissertation'. It's kind of like art - I know it when I see it. In general it should:

Writing and defending a dissertation has several purposes:

When laying out a disseration, you should think about telling a unified, complete story about the problem solution. You don't have to solve every aspect of the problem. It's o.k. to have loose ends - as long as the loose ends are not central to the theme of the dissertation and as long as you identify the loose ends and address the reasons for not pursueing them. It's o.k. to admit that time does not permit addressing something.
You should give reasons why every decision was made along the way. The more central the decision is to the main theme of the dissertation, the more justification you need for the decision. It's o.k. to do something because "that's the way it's done in the literature" as long as your research does not depend heavily on that decision.

Research proposal

A research proposal should contain the following:

Writing the Disseration or Thesis

Write. Write. Write.

I expect you to submit your Ph.D. research as a journal article or SIGGRAPH paper; a master's thesis should at least be submitted to a conference. It would be good for Ph.D. students to submit to a conference early on and the journal/SIGGRAPH article at the time you are wrapping up. The Department expects Ph.D. students to have some significant publication on their resume by the time they graduate and, of course, it benefits you when hunting for a job. (As an aside, publications are the main way I get some benefit from spending time advising you; I should be second author on any paper you write about stuff that I advise you on.)

Write technical reports. If well chosen these can be included as chapters in the dissertation. They help hone your writing skills and make you do literature searches that will have to be done for the disseration anyway. They help to lay chronological claim to intellectual territory which may be necessary if others are working in the same area. They also help to pad out your vita.

Submit conference papers. These have the same advantages as technical reports plus they give you feedback about your work from outside sources and they help to get you name known. Conferences also tend to give you feedmack in a timely manner (as opoosed to journal articles).

Of course, submit journal articles if appropriate. However, because of the long turnaround time (a year is not uncommon), any feedback you get will probably not arrive in time to help your disseration.

The thesis/dissertation should "tell a story". There should be a central story (solution to a problem) that is developed. There should not be any loose ends dangling, unless those loose ends are not central to the story. There are practical limitations (e.g. time constraints) to thesis/dissertation work and people understand this. You can't solve everything - but your work should be complete with respect to the problem that you've carved out for yourself.

A basic outline follows. Deviations from what is below are certainly possible and will depend on the subject matter:

For some general guidelines:

Get Elements of Style by Strunk and White. It's a thin book and has some gems of how to write. The basic premise is 'keep it simple'. If you can say it in less words, do so. Some basics, mostly from EoS, to watch for:

Here are some other guidelines - but note that some are more a matter of personal opinion and may vary depending on who's on your reading committee: Graphics terms: spelling, capitalization, hyphenation Also see Common Errors in English Usage

Reading Committee for Dissertation

For an M.S., get one other CSE faculty member. For a Ph.D., get three people on board as soon as you start writing. Most committee members don't get involved too early (other than the major advisor), but you should give them the opportunity to do so if they want.

Oncy you have a well-defined topic and have identified the work you must do to complete the thesis/dissertation, it's a good idea to have a meeting of your committee and present your plan. It's usually a good idea to do this with everyone present at one time, although this can also be done individually with the committee members. Sometimes it's good to do this more than once with the committee.


Oral Defense of the Dissertation

Prepare about an forty-five minute presentation. Start off with a 5 minute 'intellectual history' of yourself. This will be followed by a questioning period from the committee members.

I usually run defences so that outside people can hear the presentation part of the defense, but then I ask the visitors to leave during the question and answering period. This may change.

Planning for Presentations

Here are some on-line notes (not mine). Here's another one. Some things to think about:

Resources


Journals

What I consider to be the mainstream computer graphics journals in generally a priority order: A few that are related, but special-purpose: More geometric modeling and computer aided design than mainstream graphics: More image processing:

Conferences



Schools

  • Jobs (starting to keep track of last time I heard they were looking)


    Culture

    Intellectual pursuits aside, there is life outside of OSU. If you are from out-of-town or out-of-the-country, you may not know much about the area. Here are some things of interest to check out.

    Rick Parent
    Last updated: Monday, 23-May-2011 17:03:28 EDT