Choosing a Specific Research Problem

The first steps you should take in your search for research dollars are to choose a research problem, formulate a sound hypothesis(es), and define the research or research training program you are interested in pursuing.

Young researchers in particular must address an abundance of questions in choosing the research problem they want to tackle. Foremost among these are:

• Is the research question important?
• Will the research yield a novel or important observation (i.e., foundation for further research)?

C. Ronald Kahn, MD has written a useful, pithy treatise on the subject, titled "Picking a Research Problem - The Critical Decision", to which young researchers are referred for help in identifying their research problem of choice. Published in the May 26, 1994 issue of The New England Journal of Medicine, it is also available online to Journal subscribers.

Reviewing the funded research of other surgical oncologists will help you formulate a specific research problem. It will provide you with an idea of what is already being studied in surgical oncology and where gaps in the research exist that offer opportunities for novel or sustained research. For a look at currently funded research, search existing, online federal research portfolios. Particularly useful portfolios include NCI's online Cancer Research Portfolio, where you can search a database of current NCI grants and their corresponding clinical trials, and the CRISP database, which contains data on studies funded by the NIH.

Online literature searches such as those you can perform using PubMed will give you an idea of what has already been established in your area of research, and what further studies are needed. PubMed provides abstracts of published research findings. CTEP's State of the Science (SOTS) Web site may give you an idea of where the cutting-edge lies in your area of interest.

Additionally, you should consult NCI's Progress Review Group (PRG) reports, which highlight exciting areas of research for specific cancer sites. It is likely your grant application will garner special attention at NCI if it cites a PRG report to support the importance of your proposed research. Perusing cancer trials currently underway may also clarify your research objectives. You can search trials using the Physician Data Query (PDQ), NCI's comprehensive cancer database, or ClinicalTrials.gov.

Finally, a visit to NIH's Recently Cleared Concepts for RFAs and PAs will inform you of ideas developed by NIH that will likely turn into RFAs or PAs. Some concepts require clearance by the NCI Board of Advisors; most do not (e.g., renewals). By clicking on a recently cleared concept, an archived cleared concept, or a concept not requiring clearance, you can view the title, contact information, and objective of the proposed research. The Concepts listings provide a good tool for tracking seminal research initiatives, particularly newly developed concepts.

To map out and define your research program, you must ask yourself:

• What are your research interests?
• What disease or disease subspecialty do you want to study?
• What patient population?
• What type of research do you want to pursue - e.g., prevention, early detection and diagnosis, treatment, epidemiology, health services delivery, outcomes and survivorship issues?

Other, broader considerations to keep in mind in defining your proposed project include the following:

• Short- and long-term goals (5- and 10-year goals)
• Family imperatives
• Board requirements
• Institutional commitments
• Personal research skills
• Available resources, including mentors

Developing your hypotheses

Critical to the definition of your research program is the development of clear, specific hypotheses. Your hypothesis(es) can make or break your grant application: it is the single most important criterion for the successful evaluation of your application.

What makes a good hypothesis? A hypothesis should be:

• Highly specific
• Founded on previous research findings
• Relevant to the extent that failure of your results to support your hypothesis still adds value to the body of research. (You should, however, make sure that a large amount of negative data do not already exist for your hypothesis.)
• Somewhat new, in that it is not redundant to other, existing studies.
• Testable in a practicable way using the methods you propose.

To find examples of specific hypotheses, read through the opening and closing sections of peer-reviewed journal articles addressing research problems related to your own proposed problem. These sections illustrate how investigators justify their research to the reader by presenting a mini-review of the research literature so far published on their subject. This selective review points to gaps or trends in the research that require further investigation and that will be addressed by the proposed research.

The links listed throughout the discussion above should provide you with some foundation on which to develop good hypotheses.