StATS: Writing a research grant (September 13, 1999)
Dear Professor Mean, I'm writing a research grant to look at the impact of managed care on the care of children with chronic epilepsy. How do I structure the grant so I'm guaranteed to get funding? -- Ambitious Ann
You're writing a research grant, and you want Professor Mean to guarantee your funding? I can't do that, but if you mention my name at McDonald's you can get 10% off on your order of french fries.
It's tricky to provide advice on how to write a grant, because each research problem is different and each granting agency has its own guidelines. I can outline a few of the things that I look for when I am asked to help write or revise a grant application.
First, you need to have a theme, which I also refer to as a primary objective. Your primary objective needs to be precise, prominent, and persistent. Your primary objective will serve to unify the writing of your proposal.
Second, you need to outline your research plan. You should specify what you plan to do, who you plan to do it to, how you plan to evaluate your work, and why you are doing this research.
Third, you need to make your proposal easy to read. An abstract on the first page and a summary at the end provides organization to your proposal. Make your proposal skimmable by highlighting important points with bullets, numbered lists, and tables. Keep your proposal simple, as much as is possible for innovative research, by using analogies and avoiding OMAs (Obscure Medical Acronyms).
Fourth, write from the perspective of the reviewer. Your reviewers want information about the benefits of your research. They want reassurance that you are qualified to perform this research and that you have plans to handle unexpected problems. They also want to see consistency between your work and the interests and priorities of the granting agency.
There are four things you should think about when writing a research grant
Before I discuss each of these points, I need to outline some general categories of research.
Three general categories of research
There are many different types of research, so it is hard to describe what your plan should look like. I place research into three general categories:
. These categories aren't perfect, but help guide further discussion about writing grants.
Evaluating a treatment
Some research evaluates an active treatment or intervention for some or all of your research subjects. You will intervene throug the use of medication, surgery, or some other typoe of medical therapy. For this type of research, be sure to provide a detailed description of your treatment or intervention. Be specific. For example, we plan to hold offer a comprehensive weight loss program that includes education about diet, opportunities for exercise, and group counseling sessions.
Evaluating an exposure
Other research may involve an exposure rather than an intervention. The distinction is that an exposure is not under the direct control of the researcher. An example of exposure is breast feeding, which is a choice made by the individual mothers rather than the researcher.
Describing a population
Still other research studies are descriptive; they characterize individual variables and/or relationships among these variables. There are no treatments or exposures in this type of research. An example would be a description of pediatric patients who visit an emergency department. For this type of research, you would be interested in the average age, the proportion who have insurance, and the correlation between waiting time and staffing levels.
1. Theme: Develop a primary objective.
Your proposal should have a unifying theme. I will refer to this theme as a "primary objective" though some granting agencies may prefer to use nouns like aim, hypothesis, or goal in place of objective. It's important to understand the language that is commonly used by the researchers in your area. When they get bored, they invent criticisms like "this researcher is naive because he/she confused the aims with the objectives." These are the same people who argue about how toilet paper rolls should be hung in your bathroom.
Whatever you decide to call it, you must have a primary objective. It provides a unifying theme to your research proposal. Your primary objective should be:
You should explain your primary objective briefly and succinctly. Use a single sentence, if possible. You don't want to ramble. Also be sure to use clear and unambiguous language in your primary objective.
You should highlight your primary objective so that it is obvious even to someone who just skims your proposal. This is not the time to be subtle. If the grant format allows it, mention your primary objective on the very first page of your proposal. Otherwise, draw attention to your primary objective by using a text box, shading, or bold formatting. Use language like "the primary objective of this study is" that identifies your theme unmistakably to your reviewers.
You should refer constantly to your primary objective throughout your proposal. Repeat your primary objective at every possible opportunity. You can do this implicitly by paraphrasing your primary objective or you can directly restate your primary objective in part or in whole. Repetition will reinforce your message and will force you to keep your proposal focussed.
Make sure that you have a primary objective and that you gauge all the material in your proposal against this. Revise or remove any material that is inconsistent with or unrelated to your primary objective.
2. Blueprint: Outline the plan for your research.
State very early in your proposal what your research plan is. If the format allows it, give a brief overview of your research plan on the first page of your proposal. The plan should address:
I will illustrate these points with an imaginary research study where we develop an expensive seminar to convince people that they would be happier if they gave us all their money.
What you plan to do.
In an intervention study, you should describe in detail what that intervention is. You should also describe how you will randomize, and what efforts will be made to blind the intervention. Not all interventions can be totally blinded, of course, but you should keep the allocation to the treatment or control group hidden until the patient has agreed to participate and the recruiting physician has verified that the patient is eligible.
For an exposure study, you should focus on efforts to make your control group as similar as possible to the exposure group. Describe any efforts to match, for example, or list potential confounding variables and explain how you will adjust for them.
For both an intervention study or an exposure study, you should also describe any outcome measures that you plan to collect. Outcome variables are those variables that you believe that the treatment or exposure might influence. These measures should be specific, reliable, and valid.
If your intervention is supposed to make your patients poor and happy, you need to provide an operational definition of poverty and use a survey instrument that has been shown by other researchers to be a valid and reliable indicator of happiness.
If you have multiple outcome measures, you should list all of them in a table. If some of your measures are more important than others, indicate this priority by designating primary and secondary outcome measures.
In a descriptive study, you should focus on measurement. List all the variables that you plan to measure. If any variables are difficult to measure, explain how you will measure them.
Who you plan to do it to.
Describe the subjects that you plan to study. Detail what type of subjects you will include in your research and what type of subjects you will exclude. For example, all our patients must be rich and gullible. We will exclude any reporters and law enforcement officers from our study.
Document how many of these subjects you will see in a year and estimate how many of them will agree to participate in your research. Also estimate the proportion of subjects who will drop out during the study. You may not have a very good idea about these estimates, but you still provide your best guess so the reviewers know that you are aware of the issues involving volunteers and dropouts.
How you plan to evaluate your work.
Evaluation usually implies some type of comparison, especially for intervention or exposure studies. This can be a "before and after" comparison (before we intervened, our patients were rich and miserable, now they are poor and happy). It can also be an internal comparison (those patients who attended only one session of our intervention were still rich and miserable, but those who attended all four are now poor and happy). Or it can be an external comparison (those patients who attended our sessions were poorer and happier than a control group who did not attend our sessions).
When your research involves comparison, you should specify how much of a difference you consider clinically relevant. This information will help you justify the sample size for your experiment.
Not all research implies comparison. In most descriptive studies, for example, the focus is on statistical estimates such as means and correlations. In this case, you should specify how precise you want those statistical estimates to be. That precision will help you justify the sample soze for your experiment.
Why you plan to do it.
You need to provide a rationale for your research plan. Explain what you will learn from this research and who will it help. This is an easy thing to ignore, because it is so obvious, at least to you. Some of the "why" will already be in your literature review, but repeat and expand on your rationale when you outline your research plan.
3. Legibility: Make your proposal easy to read.
Reviewers will spend more time reading a proposal that they can read quickly and easily. Less is more. It's better to have a 6,000 word proposal that is read twice than an 8,000 word proposal that is read once. So say less, but say it in a way that will invite the reviewers to read it a second time. You do this by making your proposal:
An organized presentation is easier to read. If the grant guidelines allow for it, include an abstract at the start of your proposal and a summary at the end. The abstract and summary should discuss your theme and your blueprint. If the reviewer only reads the first page of your proposal or only the last page, make sure that the reviewer would have a good idea of what you want to do.
Your grant should follow a logical progression. When you are reviewing the literature, for example, start with the general and move to the specific. This sets up the "big picture" and places your proposal in the approprate context of the big picture.
When you are addressing a sequence of topics, use numbers to track the progress of your sequence. The "one, two, three" provides anchors of one, two, three, etc. that allow your reviewer to jump from point to point.
Don't write your grant proposal like a murder mystery. You don't want to keep your reviewers guessing about what you plan to do while you summarize the literature and all your previous research. Also use a logical and predictable sequence (such as general to specific) to make your proposal predictable. In short, you want to avoid any suspense or surprises.
You want to make it easy for reviewers to skim your proposal. Most reviewers can't devote the amount of time to reviewing your proposal that you (or they ) would like to. When you make your grant skimmable, you encourage the reviewer to 'hit the high points".
Use a topic sentence at the start of each paragraph and make an obvious transition from one topic sentence to the next topic sentence. This allows the reviewer to review your proposal by reading the first sentence in each paragraph.
Space improves skimmability. Use a large readable font. Include a lot of white space.
Use section titles liberally. If the grant guidelines include specific titles, use those title verbatim and in the same order as the guideline suggests.
Use tables, numbered lists and/or bullets to highlight key information.
Don't cram. Cramming is the enemy of skimming. I understand that the temptation to cram is strong. You want to fit in a bit of extra information. So you lower the font size from 12 points to 11.5 and then 11. You fiddle with the margins. You move some required information into the appendix. You combine short paragraphs into long paragraphs. You take out a blank line here and a blank line there. You take out some of the subheadings. And you end up with a proposal that is dense. You're proud of yourself because you got an extra 2,000 words into your proposal. Don't be so proud. In order to squeeze in a few things, you've aggravated the people who will decide the fate of your proposal.
Don't assume that the reviewer knows as much about your area of research as you do. Avoid jargon and minimize the number of acronyms that you use. If you have to use acronyms, be sure to define them at their first appearance. Also, keep the acronyms out of your abstract and out of your research hypotheses. If you can't avoid using a large number of acronyms, put in a table that lists all of them.
You should adopt a tutorial approach to the new and unusual aspects of your research. Take advantage of any analogies about your work. Although it may appear as oversimplification, you should remember that no one understands your research as well as you do. If your reviewers don't uderstand your research, they will have a hard time supporting it. We all have a tendency to mistrust those things that we don't understand. Make sure your reviewers understand your science.
Your writing style can also help simplify your proposal. Use short sentences. You should, for example, split any sentence into two or more sentences whenever you see the word "and" appearing more than once. Also use short paragraphs. Not only does this improve skimmability (see above), it also makes it easier for the reviewer to stay on track. Keep your nouns and verbs close to one another. A sentence with a lot intervening words between the subject and predicate unnecessarily causes confusion. With the subject and predicate close together, your reader will better understand your material.
4. Perspective: Put yourself in the shoes of the reviewer.
Your motivation for seeking a grant will rarely be the same as the motivation of the person who is giving the grant. Try to understand what your reviewers want and give it to them.
Reviewers are looking for four things:
One important motivator for reviewers is generalizability. If the work you plan has implications for a lot of patients beyond those that you plan to study, then be sure to brag about this. You are providing leverage; the money spent on your proposal will go a long way. For example, if you are examining the coping skills of children with chronic arthritis, mention how this work will provide direction for coping skills of children with other chronic diseases. If you are examining barriers to health care among Spanish speaking patients, emphasize how this will serve as a model for other language barriers.
The reviewer wants to understand the benefits of your work. This is not the time to be dispassionate. Let your enthusiasm for your research shine through. Advocate the value of your work through the severity of the illness and/or the number of people affected. Explain how your work fits into "the big picture."
Although you need to be an advocate, you should avoid sounding close-minded. Too strong an emphasis can give the impression that you already know the answer and you won't let any data get in the way. Avoid any overtly partisan language that might lead the reviewers to believe you are biased.
Part of your advocacy will be stating how your approach is the best. If you can't explain why your approach is the best, you shouldn't be applying for the grant. That doesn't mean that you shouldn't acknowledge the shortcomings of your approach. You should, so the reviewers appreciate your realism. Also you need to acknowledge other approaches and compare their strengths and weaknesses to your approach.
Being an advocate without sounding partisan may sound contradictory, but you should simply avoid any extremes. Either a totally dispassionate presentation or a fanatical zealotry will turn off your reviewers.
The reviewers also want reassurance. They want to feel confident the money will be spent well and there will be a good return on the agency's investment. Some things that reassure your reviewers are:
You can also build confidence in your proposal by presenting a realistic plan and budget and demonstrating care and attention to details.
If you want to be successful at writing grants, you need to understand the agency that you are applying to. What are their interests and priorities? Write your proposal to show consistency with these interests and priorities.
This is easy to do. Every agency will have guidelines for grant writers that will tell you what they are interested in. If you are like me, you don't want to bother reading these guidelines. It's like using software without reading the documentation. But reading the guidelines is very important.
Some things you want to look for in the guidelines are:
Here are some excerpts from a grant guideline.
1. Purpose: invite applications for pilot studies to explore interventions for prevention of obesity in high risk individuals or populations. Applications incorporating unique cultural or social features specific for women or for special populations defined by race/ethnicity and/or socioeconomic status are encouraged.
2. Eligibility: Applications may be submitted by domestic and foreign, for-profit and non-profit organizaitons, public and private, such as universities, colleges, hospitals, laboratories, units of State and local governments, and eligible agencies of the Federal Government.
3. Deadlines: Letter of Intent Receipt Date: March 26, 1999 Application Receipt Date: April 27, 1999
4. Money: grants will be limited to $125,000 in direct costs in the first year and a maximum of $375,000 in direct costs over a three year period.
5. Start date and duration: The anticipated award date is September 30, 1999. It is intended that the three years of funding provided for these pilot studies will allow 18-24 months of treatment and/or follow-up to be included in the proposed study designs.
Also listed in this grant are three areas that you application should address: Applications should address: the content of the intervention (e.g., relative focus on aspects of diet, physical activity, both, or other factors), the setting of the intervention (e.g., in health care settings, community groups, social groups, home, school), and the method of intervention delivery (e.g., individual, group, computer-aided, internet-based, mass media). Novel or innovative aspects and the rationale for their use should be highlighted.
This is something you should tape to the side of your computer monitor as you write your grant. Address each one of these points in detail and be sure to address them in the order in which they are stated.
There are other things you can and should do beyond reading the grant guidelines. Many granting agencies have a program official who is available to discuss your proposal. Take advantage of this person's expertise.
You should also visit the granting agency's web site. Look for a mission statement and try to show how your grant will help that agency achieve its mission. For example the mission statement for NIOSH (National Institute for Occupational Safety and Health: www.cdc.gov/niosh) is: Delivering on the Nation's promise: Safety and health at work for all people through research and prevention. If you are applying for a NIOSH grant, be sure to adress any aspects of your work that relate to safety, health, work, and prevention.
The web site might also list other information that you can draw on. NIOSH has a page discussing NORA. National Occupational Research Agenda: 21 priorities for the 21st century. Be sure to show which of the 21 priorities your grant will address.
It's difficult to provide a good example for two reasons. First, I do not have the room or the energy to reproduce a 25 page proposal. Second, it is easier for me to tell you how to write a good research grant than write one myself.
I'm hoping to get some examples of successful research grants that I can post online. If you have a research grant proposal that got funded and you are willing to share an electronic version of it for these web pages, contact me at
Here are some things you should be thinking about as you write your research grant.
Although it is targeted more towards research publications, the Mark Plonsky web page has a lot of useful and practical guidance. The article by Gotch and Stebbing (2002) is a brief list of useful tips.
This page was written by Steve Simon while working at Children's Mercy Hospital. Although I do not hold the copyright for this material, I am reproducing it here as a service, as it is no longer available on the Children's Mercy Hospital website. Need more information? I have a page with general help resources. You can also browse for pages similar to this one at Category: Grant writing.