Theses and Dissertations
 Harry Delugach

Should I do an M.S. thesis? Should I get a Ph.D.?
Typical Timeline
Thesis or dissertation advisor
Thesis or dissertation advisory committee
Thesis or dissertation proposal
Thesis or dissertation itself
Mistakes to avoid

Much of this material applies to both a Master's thesis and a Ph.D. dissertation. Since a dissertation is often called a "doctoral thesis", I'll use the term "thesis" to include both. I'll try to use "master's thesis" if there's a possible ambiguity.

For a good discussion of the purpose and content of the proposal (especially a PhD), see an excellent article by H.C. Lauer.

These are my own thoughts and advice. Students are encouraged to seek many opinions. Student are also reminded that the Graduate Handbook and the UAH Catalog contain the official rules, which may vary from program to program or from year to year. The Graduate School also has many useful resources.


Should I do an M.S. thesis? Should I get a Ph.D.?

These are important questions. Not everyone is capable of doing an adequate thesis or dissertation. A thesis for a Master's degree is intended to advance the state of knowledge in computer science, usually by applying some existing idea to a new domain, extending some existing work, or making some incremental improvement to an existing technique. A Ph.D. dissertation is a major work that advances the state of the art -- providing new fundamental ideas in computer science. A Master's thesis normally takes about three semesters, while a Ph.D. can take anywhere from six to twelve semesters.

I describe the differences between a bachelor's, master's (thesis), and doctorate in the following way:

  • During your bachelor's degree, you are presented with questions and answers that have already been developed over the years in your field of study, and you're expected to understand them.
  • During your master's thesis, you are presented witth interesting questions and are expected to develop and understand answers to the questions.
  • During your Ph.D. dissertation, you are expected to develop your own questions and then develop and understand the answers.

Unlike a bachelor's or non-thesis master's, the thesis is open-ended. In those courses-only degrees, you can plan out a program of study based on taking a certain number of courses and expect to finish in a predictable period of time. Thesis work is less structured and therefore involves more risk, both in scheduling and in knowing whether you'll finish at all. It can also be extremely rewarding: there are few times in your life when you get to follow an interesting problem wherever it may lead you, at whatever pace it demands.

(As one Ph.D. told me long ago: "Of course getting a Ph.D. is hard; if it were easy, everyone would have one and no one would place much value on it.")

If you are comfortable establishing your own work schedule, working independently, bringing up things to your advisor and coming up with your own ideas, then you may have what it takes to develop a thesis. If you are more comfortable being assigned bounded tasks in a nice sequence, then you are probably not going to be comfortable doing a master's thesis and you will definitely not be comfortable embarking on a Ph.D.


Typical time lines

The following are typical time lines for pursuing either the master's or the Ph.D. thesis/dissertation. These are guidelines only -- see the Graduate Handbook and the UAH Catalog for the official rules.

Step Master's Ph.D.
Select advisor for thesis before course work is finished before course work is finished
Finish all (or almost all) of your course work consider this time 0 (zero) consider this time 0 (zero)
Work on proposal 1 or 2 semesters 1 or more semesters
Submit proposal to committee informally (not required) at least one month before qualifying exam
Take written qualifying exam
-
soon after proposal submitted
Present your proposal (oral defense) informally (not required) at same time as qualifying exam
Meet periodically with committee optional every 6-9 months as progress is being made
Submit thesis / dissertation draft advisor's discretion advisor's discretion
Defend thesis / dissertation at least 2 weeks after draft submitted at least 3 weeks after draft submitted
Submit final ("camera ready") copies of the thesis 1 month after defense 1-3 months after defense

 


Thesis or dissertation advisor

Your advisor is a faculty member who serves as your sponsor, technical advisor, mentor and guide through the thesis process. Choosing the right advisor is probably your single most important decision in the process. Unlike an undergraduate or non-thesis degree (where the choice of even a bad professor only lasts one semester), you stay with your advisor from the beginning to the end of your thesis (exceptions are rare). I usually meet with my advisees on a weekly basis or more often.

The matching of advisor to student should be a mutual one: you should choose an advisor that you're personally comfortable with and who you think will be knowledgeable enough to guide you regarding your subject area. It's important to choose a topic that is acceptable to both the advisor and the student. If the student picks a topic that the professor isn't interested in, then the professor has little incentive or desire to really get involved. Conversely, if the professor picks a topic that the student isn't interested in, then the student will have little incentive or desire. Both situations can lead to frustration, delays and ultimately a poor thesis.


Thesis or dissertation advisory committee

In Computer Science at UAHuntsville, a master's thesis committee normally consists of three faculty members, a PhD committee consists of five faculty members. One of the faculty member is your advisor. Advisors normally help students choose the committee members. For a PhD, one committee member is from outside the CS department.

Much as in choosing an advisor, choosing committee members should be done with care. Choose faculty who you think are interested in your work, who you think will treat you fairly, and with whom you can communicate. One of the worst situations is when committee members seem unpredictable. After the first meeting or two, the student should know the basic attitude and expectations of everyone on their committee. Uncertainty is almost always a bad thing.


Thesis or dissertation proposal

For the PhD, a formal proposal is required, and is included as a part of the qualifying exam process. Most students defend their written proposal along with the oral defense of their qualifying exam. For the master's thesis, a proposal is not strictly required, nor is there a formal oral presentation of the proposal, although I strongly recommend a written proposal along with at least an informal oral presentation.

An excellent article on PhD proposals in computer science can be viewed here. I recommend it for details about what the purpose and content are supposed to be.

I suggest the following general outline for your proposal :

  • Introduction - What's the problem?
  • Background - What has already been done?
  • Research Approach - What am I going to do?
  • Task Outline and Schedule - What's my plan?

The first three sections will form the basis of the first three chapters of your eventual thesis. The purpose and content of each section is outlined below. The last section is usually relatively short and will not appear in your final product.

Introduction / Problem Statement

This section forms the problem statement for your work. You should describe some problem in computer science. The purpose of this section is to persuade your reader (a) that there really is a problem to be solved, (b) that it hasn't yet been solved, and (c) that a good (or even partial) solution to the problem would be worthwhile or beneficial in a demonstrable way. Some of this section may use quotes or material from your references, but it should focus on the problem itself, not specific solutions.

The problem statement should stand on its own and make sense to a reasonably intelligent person. After reading the problem statement, the reader should be satisified that they understand the problem you're trying to solve and decide whether they're interested in reading more.

Background / Literature Search

This section usually takes the most time in preparing a proposal. Its purpose is to show that you've "done your homework" -- i.e., that you've looked in the literature and studied what other research is relevant to solving the problem (or at least addressing it). For a master's thesis, you may have only two or three main references that you summarize, for a PhD usually more than that. Summarize the major points of each cited work (or a body of work by the same author or team). Show that you understand the previous work -- its purposes, its strengths and weaknesses, and where it fits in the overall landscape of your problem.

Research Approach

This section is the heart of the proposal; it describes what you are going to do. This section should tell what technical steps you're planning, what you expect to achieve and how it will address the problem that you identified in your problem statement. While it's usual to have bits and pieces of your approach scattered in the first two sections, this section puts them all in one place. Make clear what you think the limitations are. How will you know when you have achieved what you set out to achieve? If there are technical risks (i.e., uncertainties), identify them and what you will do to respond to them if it becomes necessary.

Task Outline and Schedule

This section is for planning and proposal purposes only. List the things you plan to do and how long you think each one will take. If there are tasks that you expect to do concurrently, show where they can overlap. Put them on a calendar and make an estimate of when you expect to finish.

Don't expect that your plan will unfold exactly as you predict. Surprises are part of the excitement of research. The real purpose of the plan is to let the committee know how much relative effort you think each of the tasks demands. This gives a good indication of whether you understand what's easy and what's difficult in your research.


The thesis/dissertation itself

There are many books and articles on the writing of a thesis or dissertation. Consider it a major work, comparable in scope and size to a short-to-medium length book.

The thesis/dissertation defense

The last major step is your thesis or dissertation defense. The word "defense" is misleading, because it implies that you are somehow being "attacked". In fact, you are defending your work with respect to any questions, misunderstandings, or criticisms that anyone might have. Most of these will be from your committee, but usually there is a question or two from people who are merely observing.

A Ph.D. defense is a public presentation of your work. Anyone may attend. You may want to invite family and friends, even if they aren't familiar with the technical details of your work.

The defense has four main parts:

Your summary presentation of the work. This should be no more than 45 minutes for a master's and no more than an hour for a Ph.D. You should summarize the important points of your work, generally following the outline of the written thesis: a problem statement, a brief background, your approach, your results and your conclusions and future work. Be sure to include a clear statement of what is your contribution: what ideas are yours and what original work did you accomplish.

Public comments from your committee. These are often done during your talk as they arise, but may also come after you're finished with the presentation. Committee members generally are free to interject questions for clarification.

Public comments from outside observers. There should be a clear invitation for questions from the audience, who are not members of the committee.

This concludes the public portion of the defense and at this point it is customary to ask everyone but the committee to leave.

Private comments from your committee. This usually involves more detailed questions, or perhaps more pointed criticism, that committee members want to explore in more depth than the public comments. This is not meant to be a "grilling" but an opportunity for members to ask you more in-depth questions.

Once these parts are completed, you'll be excused from the room while the committee decides whether to approve the work or not. Only in rare cases is the work rejected outright ("failed"). Most often, the committee agrees to certain changes that they want made before the thesis is finally submitted. Usually the advisor is charged with the responsibility for ensuring that you make these changes.


Mistakes to avoid

Each thesis and dissertation is unique, and there may be seemingly unproductive weeks or months along the way. Some research directions may not be fruitful and others may open up issues that do not fall within the scope of the original work. It's the advisor's job to help sort these out and help the student identify when they've happened. These sorts of tangents, dead ends, etc. aren't "mistakes" in the sense that they are an inevitable part of an open-ended research process.

There are, however, things that a master's or PhD student may do that will get them into trouble in various ways. These are some of the mistakes that students have made in the past.

Choosing the wrong advisor

Unlike an undergraduate or non-thesis master's degree, your overall performance and satisfaction with a thesis will depend largely on your advisor. Choose him or her wisely. Meet with them. Talk with them. Get other student's impressions and experiences. Look over their web page and check into their publications. Find out what you can about their research interests and their history in working with students. While you should certainly choose an advisor whose interests are close to yours, do not feel compelled to choose a particular one. The choice is yours.

Not heeding your advisor and/or committee

Sometimes students ignore or dismiss their advisor's suggestions. This can lead to erroneous results, incomplete theses, poorly done studies, badly formed approaches, etc. Trust your advisor! If you don't trust your advisor, then consider choosing a different one.

Of course, you don't have to do everything your advisor says. But you must pay attention to everything your advisor says; if you decide not to take suggestions, then you should consider them reasonably and rationally and provide a clear defense of your choices. This is even true when you have taken your advisor's advice: always have a justification for your decisions.

Often committee members will have different ideas about your work (that is why you have a committee). If there is a conflict between members of your committee, that is a matter for your advisor to sort out.

Not accepting criticism

You should accept the fact that your work will not be perfect! There will be flaws, mistakes and omissions, regardless of the effort made by yourself, your advisor and your committee. A lack of comments on your thesis usually means a lack of interest or lack of seriousness in evaluating it. Addressing comments will usually make it a better work. Bear in mind that errors don't just disappear on their own -- either your advisor and committee will help you find them now, or else external reviewers and potential employers will find them later.

Resist the urge to defend every one of your decisions whenever criticism is leveled. In other words, try not to react with "but this is why I did it!" when one of your decisions is criticized. On the other hand, be prepared to defend decisions that you think are right. In all cases, listen to the criticism and pay attention to its substance. This is related to the previous point about heeding your advisor and committee.

Sticking to an arbitrary timetable

Students sometimes embark on a master's or PhD with a fixed idea of when they will finish. ("I plan on spending spring, summer and then finish in the fall.") Your plan may or may not be realistic. Remember that the completion of a thesis is not measured on the calendar, but is measured by the thesis's content and quality. If it's not finished by your arbitrary deadline, then you'll have to spend more time than you thought. Don't consider this a failure or shortcoming on your part! No one "blames" you or is prejudiced by your taking longer than you originally planned. Since your thesis topic drives you toward your completion, sometimes it may lead you into new areas that you never imagined would be relevant, and sometimes (often?) things take longer than you originally thought.

Do not blame your advisor or committee either. No one wants to delay you without a reason, and everyone wants you to produce the best thesis you can. See above for how to handle criticism when it occurs.

Do not try to rush a thesis to defense before it is ready. At best, it will likely take you months more editing anyway, and at worst your whole degree could be in jeopardy.

Defending a thesis "blind"

Do not walk into a thesis or dissertation defense unless you know what every member of the committee (including your advisor) already thinks about the work you're defending. Talk to each of them before you defend. Get a sense of what they liked, what they didn't like, where they think you did well, and where they think you need to make improvements. Even strong criticism can be handled, if you know about it in advance and are prepared to address it. If you think a committee member is inclined to vote against approving your thesis, you should discuss this immediately with your advisor. If your advisor seems to have major reservations or misgivings, then DON'T DEFEND! Postpone or delay your defense until your advisor is satisfied. (See above about sticking to an arbitrary timetable.)

 

 

This page last modified Sunday, August 24, 2008