Careful planning is an essential element of all good research studies. This guide provides a useful reference to help you ensure that your study is well designed and that data can be collected effectively, with optimum utilisation of resources.

First, a clearly defined hypothesis is needed, e.g. "there is a relationship between calcium intake and bone density". The study should then be designed so as to test the hypothesis. Strictly, the hypothesis should be phrased as "there is no relationship between calcium intake and bone density" (the null hypothesis) and the study designed so as to disprove the null hypothesis. Sometimes data may be examined to see what hypotheses they suggest. However, this should only be done as a preparatory process and needs to be followed by a separate study in which a hypothesis is tested with new data.

Next, you need to decide how the hypothesis will be tested. There are several methods to choose from - click on an item in the table below to find out more about that method.

Study Method	Type of Research
Cross-sectional Survey	Describe a population, e.g. to ascertain mean nutrient intakes and compare these with Dietary Reference Values. Ascertain what proportion of the population have a particular condition, e.g. high blood pressure. Compare different groups of people with respect to various characteristics, e.g. smokers versus non-smokers. Investigate associations between two or more variables within the population, e.g. nutrient intakes and blood lipids.
Case-control Study	Subjects who have the disease of interest (cases) are compared with subjects who do not have the disease but are otherwise similar to the cases (controls)
Prospective (cohort) Study	Defined population sample is followed forward in time. Variables of interest are assessed at baseline, before development of the disease. At the end of the follow-up period, baseline measurements are compared between those who went on to develop the disease and those who remained free of the disease.
Intervention Study	Also known as a randomised controlled trial. Half of the subjects alter a specified variable, e.g. increase fruit & vegetable consumption, whilst the other half make no change. All subjects are then followed up for a defined period to ascertain whether the intervention has any effect on the outcome measures, e.g. nutrient intakes, blood lipids or risk of heart disease.

You need to calculate the number of subjects needed for the study, to ensure there is a high probability of detecting a difference, if it exists. Click here for details. Then, you need to select subjects for your study. Click here for details. It is advisable to obtain the advice of experts in the field to ensure that, as far as possible, the answers you obtain will be widely accepted as a valid test of the hypothesis. You should also consult a statistician while the study is being planned. He/she can advise you on the number of subjects needed, various details of the study and how best the results should be analysed.

If the study is concerned with food/nutrient intake, you need to decide on the most appropriate dietary survey method. There are several methods available. Click on an item in the list below for details.

• Precise weighing method
• Weighed intake record
• Food diary
• Diet history
• Diet recall
• Food intake questionnaire

The feasibility of the study needs careful consideration. How much depends on the goodwill or active co-operation of colleagues? Forgetfulness, loss of interest and changes in staff can be a problem if the study depends on active participation of a number of other persons. A pilot study (small preliminary study using similar subjects to those who will take part in the main study) is very useful in identifying other potential difficulties.

The amount of data to be collected needs to be considered. It is obviously efficient to get as much useful information as possible from the study. However, there is no point in collecting items of data that will never be used. Each item therefore needs to be looked at critically and its inclusion justified.

The study should be explained to subjects in non-technical language - people are usually willing to participate in a study if they understand its purpose and can see its potential importance. In terms of what is expected of the subjects who take part in the study, and how they are dealt with: "do unto others as you would have them do unto you". The implications of the study outcome should also be considered. For example, if there is a positive outcome will this lead to a change in current practice? Does the study test a procedure that could be followed in real life or one that could exist only for the purposes of an experiment?