Basic Statistics for Researchers
A self-paced, hands-on statistics guide in R
Welcome
Most statistics courses start with theory. This one starts with your data.
Every method here is introduced through a concrete research scenario: the kind of situation you will actually face in your PhD or your lab. You will see the R code, the output, and a plain-English explanation of what each number means and how to report it. The goal is not to make you a statistician. It is to make you a researcher who can analyse their own data, understand what the output is telling them, and write it up correctly.
Go to Before You Begin: Setup and follow the steps there. You will install R, RStudio, and all the packages the course uses. This takes about 10–15 minutes and only needs to be done once. Nothing else in the course will work until you have done this.
Who This Course Is For
This course is written for researchers who analyse their own data: PhD students, clinicians, epidemiologists, and biomedical and life-science researchers. If you work with patient records, cohort or registry data, lab measurements, or any quantitative health data and need to analyse it yourself, this is for you.
The examples are drawn from real clinical and biological datasets: patient survival, disease registries, NHANES population-health data, and genetic-epidemiology studies, so the worked analyses look like the problems you actually face. The methods themselves are general, though: a t-test, a regression model, or a survival curve works the same whether your data come from a clinic, a field study, or a lab, so researchers outside the health sciences will still find everything here directly usable.
This is a foundations course: it teaches the core methods you need for day-to-day research, taught hands-on in R. It is deliberately not an advanced-biostatistics, clinical-trial-design, or meta-analysis text. The goal is to get you confidently analysing, interpreting, and reporting your own data, and knowing which method fits which question (see the Statistical Decision Guide).
What This Course Expects From You
Not much. You need R and RStudio installed, a basic ability to run code (copy, paste, press Run), and a willingness to read output carefully. You do not need a statistics background. You do not need to be good at maths. High-school algebra is enough.
What will slow you down is skipping the output. Every session shows real R output with a table that explains each number line by line. Read those tables. That is where the understanding happens.
How the Course Is Organised
The course has four parts. Each part builds on the one before it.
| Your situation | Where to begin |
|---|---|
| New to statistics entirely | Part 1, work through in order |
| Familiar with R, new to modelling | Part 1, then Part 2 |
| Comfortable with inference, need regression | Part 2 |
| Need advanced methods (survival, causal) | Part 3 after Parts 1–2 |
| Just need to know which test to use | Part 4: Statistical Decision Guide |
When in doubt, start at Part 1 and go in order. The sessions are short enough that skimming a familiar topic costs ten minutes at most.
Part 1: Inference, Testing, and Estimation
This is the foundation. You will learn how to reason from a sample to a population, what a p-value actually means (and what it does not mean), and how to compare groups. If you have ever reported a p-value without being entirely sure what it means, this part is for you.
- Describing data: distributions, variance, and normality
- Sampling, confidence intervals, and the central limit theorem
- Hypothesis testing: the logic of evidence
- t-tests and group comparisons
- Non-parametric tests for when assumptions fail
- ANOVA for more than two groups
- Categorical data and chi-square tests
- Power and sample size planning
Part 2: Regression and Modelling
Regression is the workhorse of biomedical research. This part covers the models you will encounter most often: linear regression for continuous outcomes, logistic regression for binary ones, and models that handle repeated measures and missing data. By the end you will be able to fit a model, check its assumptions, interpret the coefficients, and write a Methods section that describes what you did.
- Correlation and association
- Simple and multiple linear regression
- Logistic regression for binary outcomes
- Model building, diagnostics, and validation
- Missing data and multiple imputation
- Mixed models for clustered and repeated-measures data
Part 3: Advanced Methods
These sessions cover methods that come up once you move beyond standard comparisons and models. They assume you are comfortable with the material in Parts 1 and 2.
- Survival analysis and time-to-event data
- Causal inference using directed acyclic graphs
- Mendelian randomisation for genetic instrumental variables
Part 4: Reference and Resources
A set of reference materials you can return to at any point.
- Statistical Decision Guide: which method to use for which question
- Clinical research methods and study design
- Help and documentation
The Statistical Decision Guide in Part 4 is a one-page reference card. If you are in the middle of an analysis and cannot remember which test to use, start there.
How Each Session Is Structured
Every session follows the same seven-part structure. Once you have been through one session you will know exactly where to look in every other.
- When do you use this? A one-paragraph scenario that puts you in the situation where this method is the right tool.
- Background The core idea in plain language, with just enough mathematics to understand what the method is doing.
- Worked examples Step-by-step R code on real datasets, with output explained line by line.
- What can go wrong The most common mistakes, misinterpretations, and ways to misuse the method, and how to avoid them.
- Exercises Three problems: one guided with hints, one semi-guided, one open-ended using your own data.
- Comprehension check Five questions with detailed answers. Use these to test yourself before moving on.
- How to report Exact language for your Methods and Results sections, with examples of correct and incorrect phrasing.
You can work through the course week by week, or jump directly to a specific method when you need it. If you are mid-analysis and unsure which method applies, the Statistical Decision Guide is the fastest route to an answer.
Course visits
