Build an analysis plan before choosing a statistical test. For a health research proposal, begin with the question, the outcome, and the way observations are collected.
Learn biostatistics with open-source software
We promote open-source software for learning and conducting biostatistics. Start with R for a code-based workflow or jamovi for a graphical interface, and keep a record of your analysis decisions.
R: learn a code-based workflow
RStudio Education: beginner resources for learning R provides starting points for installation, working with data, visualization, and reporting. R is a free, open-source language and environment for statistical computing; RStudio is an interface used to work with R.
jamovi: start with a graphical interface
jamovi: getting started introduces opening data, setting up variables, running an analysis, and saving your work. jamovi is free, open-source statistical software powered by R.
Follow a worked tutorial
Your First Descriptive Analysis in R and jamovi: A Free Biomedical Dataset
Practice with a free vitamin C and dental-growth dataset from an animal experiment. Follow the R and jamovi steps, compare your summaries with the reference tables, and learn to write a careful interpretation.
A practical first exercise
- Choose a public or synthetic dataset and define one research question.
- Check the variable types and summarize the data.
- Choose an analysis that fits the question and design, then review its assumptions.
- Save your R script or jamovi project together with notes on the settings and software version.
- Write a short interpretation that includes uncertainty and limitations.
Software supports the analysis; understanding the study design and assumptions remains essential. Learning resources checked: 10 October 2026. Read more about R on the official R Project website.
Five questions to answer first
- What is the objective? Describe a population, compare groups, estimate an association, or predict an outcome?
- What is the outcome? A measurement, a category, a count, or time until an event?
- How are observations related? Independent participants, paired measurements, repeated visits, or people clustered within clinics?
- What needs adjustment? Identify plausible confounders using subject knowledge and the study question.
- What could affect interpretation? Missing values, small groups, measurement error, selection bias, and multiple comparisons.
Prepare an analysis worksheet
For each objective, record the outcome, explanatory variables, unit of analysis, proposed summary, planned model or comparison, assumptions, and sensitivity checks. Treat this as a discussion document for your supervisor or statistician.
Describe the data before interpreting results
Make a data dictionary with variable names, definitions, units, coding, and missing-value rules. Check participant counts and inspect distributions. Keep a record of data cleaning decisions so another researcher can follow your work.
A useful practice exercise
Imagine a study comparing sleep quality scores between two groups of postgraduate students. Before selecting a test, ask whether the groups contain different people or repeated observations of the same people, how the score is measured, and whether the comparison requires adjustment. The variable name alone cannot settle these questions.
Write an interpretable results paragraph
State the analysis population, summarize the outcome, report the effect estimate and its uncertainty, and describe relevant limitations. Explain the result in the context of the original question rather than relying on a significance label alone.
Prepare for postgraduate study
Practice with a small public or synthetic dataset. Submit a short analysis plan, a reproducible script, and a results paragraph as your learning portfolio. Follow the application rules before including supplementary materials in a scholarship application.
Continue with research methods, sample size planning, or academic writing. Return to the Health Research & Graduate Funding Hub.