First question: Do you need one?

Maybe not. Good for you.

What will a grant pay for, and what do you need to do your project?

Stipend for living expenses and time off from TA/RA duties? An annual salary.
Research expenses? Computer, travel expenses, wages to hire an assistant, books, datasets, subject fees to do an experiment, costs to implement a survey?
Both?

What are the potential sources?

Internal university fellowships for time off with no duties.
RGSO and other internal awards for research expenses such as a plane ticket.
Specialized foundations, governments, especially for foreign study
Other specialized foundations, visiting post-doc and pre-doc fellowships at centers such as Harvard, Princeton, Brookings, etc.
Self-financed trips made by taking a part-time job. How risky this route is.
NSF dissertation awards (research expenses only, up to about $8k).
No single list, need to work with your advisor and others to search for potential sources.
Importance of getting familiar with the sources of funding in your field.

Why bother?

Some reasons not to:

It takes a lot of time that you could be spending directly on doing the project rather than on this preliminary step.
It imposes another lengthy time delay in the process. Typically 3-9 months before you hear.
Many projects just don’t need it.
Your project will be evaluated in the end by the output (publications), not by the inputs.

Some reasons why it might be worth it:

Sometimes bigger is better. For example, with a $3,000 research grant you could hire six undergraduates with work-study to work for you 10 hours a week for 12 months. That’s a lot of research assistance. They could help you implement a survey, code newspaper stories or other documents, or do whatever else your dissertation calls for.
Fieldwork-related projects require funding, absolutely.
It can be good training for life as a professor. Professors either get into outside grants or eventually they find that their university does not support them in their research.
It puts you through a peer-review process and you get some feedback.
Review boards tend to be major people in field. Success there is a good sign and people might remember you or be helpful in the future.
Success sends a signal to the profession about the quality of your work. Helpful in getting a job.
The time delay and the peer review by outsiders force you to sharpen your ideas considerably. This is never harmful (though it can be painful).

What does a proposal look like and how will it be evaluated?

§ Some focus on the previous accomplishments of the student, with less attention to the proposed project.

§ Different granting agencies are looking for different things. Some want to encourage the study of a certain language, some to focus on theory, some on a substantive topic, some to support students of certain backgrounds. Match the proposal to the agency. Pay attention to the criteria. Address all stated criteria explicitly.

§ Project description

o Clear statement of the theoretical idea and goals of project.

o Good command of the literature. (Demonstrated, not hinted at.)

o Good command of appropriate methods. (Note the plural. Also demonstrated, not to be guessed at.)

o Feasible research strategy, by this individual applicant. Again, this must be demonstrated.

o Endorsement by the advisor.

§ Evaluation criteria (can vary by funding agency, as noted above)

o Importance of topic, likely impact (different standard for $8,000 than for $250,000)

o Quality of the treatment (lit review, methods, writing quality, organization)

o Feasibility by this particular student

Some final thoughts

§ The bar is not so high as you might think, even at a place like NSF

§ Simultaneous submissions encouraged (not like journal articles).

§ Internal grants can be very good practice for larger later ones.

§ Explaining your project in an abstract or two-page format is hard to learn but very useful.

§ The batting average doesn’t matter. Only the number of times you get on base.

§ Go to bat early and often.