Design a research study

Taking into account the results of the literature review, and the gaps that you can fill in your own research project, you are ready to start to plan your study. Before you start, here are some resources to help with the possible methods you can consider.

Here are the major headings you need to consider in designing a research project.

1. Define the Research Question.

A research question is a question that a study is designed to answer. It guides all other subsequent tasks in the research process and at the end of the project, after data collection and analysis you should be able to answer your question.

The purpose of the research question includes: Organising the study and giving it direction and coherence; Delimiting the study, showing its boundaries; Providing a framework when you write up your research; Keeping the researcher focused; Pointing to the hypothesis, methods and data that will be needed.

Characteristics of a good question include being: Ethical; Novel; Interesting; Relevant; Feasible.

NOTE: the above information on the research question is taken from the Global Health Training Centre and you may want to look at  The Research Question through eLearning courses (you will have to register, but it is free to access)

2. Develop the hypothesis. This will be based on the gaps in knowledge you identified in your literature review. In the Research Skills Programme part 2 below, you  will find the folder Research hypothesis revisited which contains RSP_03_The research hypothesis.pdf which provides the way to develop a research hypothesis and RSP_03_Example from Research Idea to Research Hypothesis.pdf which gives two simple examples from previous students.

NOTE: For quantitative studies you will need a falsifiable research hypothesis; i.e. a hypothesis in which you specify the expected outcome in numerical terms. For qualitative studies a proper research question/topic is sufficient. Qualitative studies do not work with hypotheses.

3. Pick your study design. Please look at these: Overview of research designs RSP_04_Overview of research designs.pdf  Qualitative-research-methodology_MSF_2007.pdf will help if you are planning a qualitative study.

NOTE: For every research question/hypothesis there is an ideal study design. Please check out RSP_05 The right research design for your research question.

You might want to review the WHO book "Basic Epidemiology" by Bonita, Beaglehole and Kjellstroem (download pdf here). 

4. Select your sample to study.

NOTE: The target population is the population to whom you want to generalize your results to. Again this does not apply to qualitative research as these studies do not want to generalize results to a wider population.

5. Clearly define your sample with inclusion and exclusion criteria. If you have a complicated sample, such as a cluster sample you also need to define the “clusters” with inclusion and exclusion criteria.

6. Sampling: Please consider different sampling strategies and read RSP_06_sampling and sample size.pdf on sampling and sample size.

NOTE: Sampling is usually different in qualitative studies which want an information-rich sample. In contrast, quantitative studies usually want a representative sample using probability sampling.

7. Sample size: In a quantitative study the required sample size is directly linked to the falsifiable research hypothesis. Depending on the research hypothesis DIFFERENT formulas apply. The simplest applies when estimating the prevalence of a disease using a cross-sectional study RSP_06_sampling and sample size.pdf However, formulas change if you want to compare groups of people, and again if your sampling design is complex (e.g. cluster sampling). Many sample size calculators are available online. Please ask us if in doubt!

Qualitative studies sample until the data reaches “saturation”! So the sample size in qualitative studies cannot be calculated beforehand.

8. Make measurements of the exposure and the outcome and potential confounders. Clearly describe and define your main variables in your study.

9. Plan how to analyze the results. Please read RSP_07_Statistics with confidence.pdf

NOTE: The statistical analysis of the study has to be linked back to your research hypothesis. How will you answer your research hypothesis with statistical techniques? Similarly for qualitative studies – the analysis has to be linked back to the research question. Although, qualitative studies do not apply statistics to reach conclusions.

10. Avoid bias in sampling, measurement and confounding. Please read RSP_08_Bias.pdf The topic about bias is important when planning a study and also links to the limitations of the study in your research proposal where we expect an honest account of expected bias.

11. Consider the time frame and ethics.

NOTE: The time frame is about the feasibility of your study. Please consider the required sample size and how long it will take you to recruit the number of people.

Ethics should also reflect on the particular situation of the study and how participants might be affected by the study and how negative effects could be minimized.

12. Consider the potential limitations of the study. Consider selection bias (whether your sample represents your target population); information bias (whether the information on exposure and outcome are valid); and confounding (whether results might be distorted by other variables). Please also comment on how to minimise these biases in your study. Please read RSP_08_Bias.pdf

NOTE: Qualitative studies are not so much worried about bias – but they need to show that their results can be corroborated with other findings. This is called “triangulation”. Please read the MSF document carefully if you are proposing a qualitative study.

RESEARCH SKILLS PROGRAM PART 2 

The second part of the RSP provides you with resources that introduce quantitative and qualitative research methodology. Please use the RSP Part 2 as a support for developing your research proposal.