Skip navigation
Brigham Young University
CS Department Department of Computer Science

Tips for Graduate Writing: Proposals

Grad Student Handbook

Before you begin, make sure you review the guidelines on proposals in the Graduate Handbook. Follow the organization described there.


The introduction section should clearly lay out what the problem is and why it's important. Don't leave out that last part--plenty of work has been done solving unimportant problems.

As part of the introduction, you may also have to explain some basic concepts to the reader who may not be an expert in your particular area of CS. How much is enough? For me, the answer is, "enough that the reader can understand and appreciate your thesis statemet when they get to it".

Literature Review

Your proposal should also review the relevant (emphasis on relevant) literature that pertains to your intended research. You will also include an annotated bibliography, so don't feel you have to describe every paper you've ever read on the subject. Don't write one paragraph saying, "FamousGuy did this..." and another saying, "BigCheese did that..." Instead try to tell a story: "FamousGuy first proposed a system for doing such a such using Method A. BigCheese built on that by adding Methods B and C. Kahuna pointed out the weakness in Method B and proposed instead to use Method D", and so on. Each of these can be expanded to a paragraph or two if necessary. Another approach is to use a thematic organization of the field: "Some researchers have used Method A (FamousGuy, Big Cheese, Kahuna), which has the advantage of blah, blah but requires yada, yada. Others have used Method E (OldProf, ViciousRival), which addresses these shortcomings but instead requires blah." Even if this does devolve to "Person A did Method B", at least try to organize it for the reader.

When writing the literature review (or background sections for papers), keep in mind the following three goals:

  1. You want the reader to be convinced that you've done your homework and are familiar with existing literature. In other words, your stuff is new and placed in appropriate context.
  2. You want to pay homage to those who came before you. Your literature review should give credit where credit is due and recite appropriately the litany of past work. (This is most important for papers--the person whose work you leave out is almost certainly by some cosmic law going to be asked to review your paper. I say this from personal experience.)
  3. You want to educate your reader. Use the literature review to teach them what they need to know in

Your Thesis Statement

Webster's defines a "thesis" as "a long essay or dissertation involving personal research, written by a candidate for a college degree". However, it also defines it as "a statement or theory that is put forward as a premise to be maintained or proved". The latter is the correct way to thing about your thesis statement: what you intend to show, prove, or demonstrate by your research.

It is common for "thesis statement" to be misinterpreted to mean a statement of what you intent to do for your thesis. It is not that! When you find yourself describing what you will do rather than what you will show, you've lost track.

Here's another way to think about a thesis statement. Most papers you read include a background or literature review section that includes things like, "FamousGuy (1960) showed/demonstrated/proposed ____." Some day, you'll be the one whose work is being cited. How do you want them to fill in this blank? That's your thesis statement.


The goal of the methods section is to lay out a plan of attack on the problem sufficient to convince your committee that it has a chance of succeeding. This is to avoid the "Here's a great problem, but I don't have a clue how to solve it" proposal.

Obviously, though, many of the details of your methods are still to be worked out--that's why it's a proposal and not a full thesis at this stage. Don't worry if the exact methods aren't certain just yet, but at least have an idea how to approach it.

Means of Evaluation

This is one of the most overlooked parts of a thesis. Besides just implementing a solution to the problem you're working on, you have to be able to assess how well that solution worked. Is it better than previous methods? Faster? Cheaper? More usable? Remember your thesis statement: what do you intend to show? Not just do, but show. How do you plan to show it? Implementing something is usually part of it, but once implemented, how do you demonstrate what you intended to?

Last modified: April 30, 2007. Maintained by Bryan Morse.

Copyright © 1994-2005. Brigham Young University. All Rights Reserved. XHTML CSS 508