There are a number of well-established steps to undertaking a research study in the pre-clinical or clinical sciences. Successful research requires a systematic research process.
You can download a pdf of this page here: Before you start
- Formulate your research question or series of questions. This is possibly the most important part of your study. Using the PICO (Population, Intervention, Comparison and Outcome) format will help you formulate an answerable research question.
- Select the best study design to answer your question. The most common study designs are randomised controlled trials or observational studies, which encompass cohort, case-control or cross sectional (survey) studies. Each of these has particular strengths and weaknesses. See the SPIRIT and CONSORT statements for randomised controlled trials; STROBE statement for observational studies. For a comprehensive list see the equator network http://www.equator-network.org/. If you are not confident in this area, talk to an epidemiologist or biostatistician BEFORE you select a design for your project. Download a pdf guide to types of study here: Design: types of study
- Plan your data collection and analysis. You need to plan in advance how you will collect, store, and analyse your data to answer the research question. Download a guide to data collection here: Data_collection_and_spreadsheet_management. For a cheat sheet on statistical methods, see: Statistics cheat sheet.
- Estimate the sample size you will need to have a pre-determined chance of accepting or rejecting your hypothesis. For example, it is conventional to set the power of a study at 80%, or an 80% chance of detecting a difference between groups, if there is one. You also want to have a low chance of making a ‘false discovery’, conventionally set at 5%. These concepts are known as ‘Type II’ and ‘Type I’ errors. Estimate your sample size based on the primary outcome for your study. See Sampling and sample size calculation.
- Consider the ethical implications of your study. What is the potential for benefit or harm to your study participants and the wider community?
- Write a detailed study protocol. This will be the ‘Bible’ which details all aspects of your study. The protocol will need to be submitted to your institutional ethics committee for approval. (You may have to adapt your protocol to conform to a format required by the ethics committee).
- Conduct your study.
- Analyse and write up your results. It is strongly recommended that you document all computer code to make sure that your results are reproducible. Write your manuscript. For a concise but comprehensive guide to statistical reporting see SAMPL guidelines.
- Submit your manuscript to a suitable journal, aiming to reach your target audience.
- Almost always, even if your manuscript is accepted you will need to address the reviewer’s comments. Address these as fully as you possibly can. Remember that reviewers work for nothing, and their comments will only improve your communication.
- Your paper is accepted! Congratulations and well done.