Sharp: a method for developing a research proposal — cutting edge or blunt tool?
Trudy Harpham
South Bank University, London, UK

S A J Clin Nutr 2000 February Vol 13 No 1

Supplement
Researchers in the south, particularly South Africa where investment in research is relatively high, are increasingly well positioned to compete for international research funds. Many northern-based research funds (e.g. Wellcome Trust UK, Department for International Development UK) are now actively seeking proposals which emerge directly from the south rather than being ‘ventriloquised’ through an academic institution in the north. This is a significant change and requires researchers in the south to gain and use the capacity to write proposals which will stand up to fierce international competition. For example, researchers in the UK who are judged by the government in part by their performance in raising research funds are used to writing several proposals a year and typically discuss their success rate with colleagues. In South Africa, it appears that few researchers have had experience of applying competitively for international research funds. This was partly behind the International Union of Nutritional Sciences (IUNS) Urban Nutrition Action workshop’s decision to mount a training workshop on both research and intervention project proposal design. In terms of research proposal design, any number of guidelines could have been used. The classic text, Proposals that Work, 1 is an example of a guideline that can be used by almost any discipline. However, the fact that SHARP(Structured Holistic Approach for Research Planning) had been published in the Food and Nutrition Bulletin in 1998 2 and that one of the authors was a facilitator at the Durban workshop in 1999 determined the choice of guidelines.

A particular strength of SHARPis the enforced development of a causal model during which all important variables and their relationship to the central hypothesis are defined. This ensures that existing literature is used and helps researchers to avoid ‘re-inventing the wheel’. However, one problem is that much effort can go into forming a causal model de novo yet, particularly in a sub-field like urban nutrition, there is a finite number of relevant key variables, therefore most models end up looking very similar. Indeed, the one produced at the March 1998 meeting, described in the following pages, bore remarkable resemblance to the example provided in the original article by Gross et al. 2 One therefore questions the need for developing specific, additional causal models. Future researchers are advised to make use of Gross et al.’s model 2 and to elaborate it accordingly.

Another problem of SHARP is the requirement for hypotheses and its related steps that suggest it is only suitable for quantitative research (e.g. sampling procedures, statistical methods). This assumption of quantitative research is not stated by Gross et al. 2 Qualitative research on, for example, the difference between rural and urban population’s construction of the meaning of food, does not readily lend itself to the application of SHARP. Although it might be argued that such qualitative research would often be ‘pilot’ or ‘ preliminary’ research leading into quantitative research, many social scientists would argue against positivist hegemony which insists on hypotheses. Sound qualitative research has clear objectives, but does not necessarily pose causal pathways. However, use of SHARP as an aide-memoire for all the necessary sub-sections of a proposal is recommended. It should be borne in mind that it was developed for students and, therefore, the process is as important as the outcome. What outcomes does SHARP tend to produce? One observation is that, like ZOPP, considered elsewhere in this issue, the root causes of malnutrition and related problems are placed at the bottom of the causal pathway tree. One can argue that this increases the significance of factors such as poverty and inequalities (by placing them at the ‘root of the tree’), but in practice they tend to be ignored as the tendency is to focus on intermediate variables nearer to the key outcome variable at the top of the tree (e.g. nutritional status of preschool children). This might discourage research on the structural causes of nutritional problems, which in turn may diminish interst in the nutritional field by other disciplines.

SHARP can also blunt the action or follow-up in key areas of interest. For example, a fascinating research problem is the coexistence of maternal obesity and infant/child malnutrition within black households in South Africa. If one imagines applying SHARP to this problem, one can envisage a traditional epidemiological-type study emerging which fails to capture the dynamics of a subtle process which lies behind this picture. Exploratory, qualitative research might be more appropriate in this instance.

So, is SHARPa cutting edge or a blunt tool? For researchers at an early stage in their career with an epidemiological (or at least quantitative) bent it can certainly sharpen their research proposals. For multidisciplinary work or for qualitative research it may be too constraining; it risks inhibiting cutting edge nutrition research aimed at illuminating the processes which produce nutrition inequalities.

1. Locke L, Spirduso W, Silverman S. Proposals that Work. London: Sage, 1987.
2. Gross R, Karyadi D, Sastroamidjojo S, Schultink W. Guidelines for the development of research proposals following a structured, holistic approach for a research poposal (SHARP), Food and Nutrition Bulletin 1998; 19: 268-282.

Last updated: 17-Feb-2004