R NovaR
Statistics with R

Learn R by running it.

Thirty courses take you from your first mean() to mixed models, forecasting, clustering and random forests — in R, the language statisticians actually use. Every lesson comes with a lab that runs real R in your browser: you write the code, you see the console and the plot, the server tells you whether the answer is right. Nothing to install.

Start the free course See all 30 courses

The first course is free in full — lessons, labs, final exam and the certificate. No card, no trial clock.

🧪
Real R, in the page

Not a quiz about R — R itself, running in your browser. Write t.test(x ~ group, data = df), read the output, get told if your answer is right. 237 labs so far.

🎓
University-level, not tips

Assumptions, when a test breaks, what a p-value really says. The maths is explained, then done — with the exact R function and its exact output.

📄
A certificate per course

Pass the final exam and get a certificate with a public verification link and a PDF — one for every course you finish.

The honest part

Assumptions first. Then the function.

R will happily run a t-test on data that violates every assumption the test makes, and print a p-value to four decimals. These courses teach the assumption, how to check it in R, and what to do when it fails — before the function call. Base R where it is enough, dplyr and ggplot2 where they genuinely help, and a clear note whenever a method needs a package the browser lab cannot load, so you know what to run in RStudio.

Curriculum

Thirty courses, in order

30 courses · 240 lessons · one certificate each

🧮 R Basics
The console, vectors and functions: the three things every R script is made of.
Free, in full
🗂️ Data Structures
Types, factors, lists and data frames: how R stores data and how you get at it.
8 lessons · Beginner
🔁 R Programming
Conditions, loops, functions and the apply family: turning one-off console work into code that runs on its own.
8 lessons · Beginner
🧹 Data Import and Cleaning
Reading files, fixing types, dates and missing values, and reshaping until every row is one observation.
8 lessons · Beginner
🔗 dplyr and tidyr
Six verbs, the pipe, joins and reshaping — the grammar behind almost every R data pipeline.
8 lessons · Beginner
📈 ggplot2
The grammar of graphics: build any plot from data, aesthetics, geoms, scales, facets and themes — and know what it can hide.
8 lessons · Beginner
📊 Descriptive Statistics
Centre, spread, shape, quantiles, outliers and grouped summaries — and why quantile() disagrees with Excel.
8 lessons · Intermediate
🎲 Probability Distributions
The d/p/q/r functions for normal, binomial, Poisson, exponential, uniform, t, chi-square and F — plus simulation and normality checks.
8 lessons · Intermediate
🎲 Sampling and the Central Limit Theorem
Why one sample can stand in for a population, how far its mean can be off, and what shrinks that error.
8 lessons · Intermediate
📏 Confidence Intervals
Turning a standard error into a range for the mean, a proportion or a difference — and reading that range correctly.
8 lessons · Intermediate
⚖️ Hypothesis Testing
Null and alternative, the p-value and its three wrong readings, errors, power, multiple comparisons and p-hacking — every claim checked by simulation.
8 lessons · Intermediate
🔬 t-Tests
One-sample, Welch, Student's and paired t-tests, their assumptions, Cohen's d, confidence intervals and what to do when the assumptions fail.
8 lessons · Intermediate
📐 ANOVA
One-way and two-way aov, the F statistic by hand, assumption checks, Tukey HSD, interactions, eta-squared and the Kruskal-Wallis fallback.
8 lessons · Intermediate
🔲 Chi-Square Tests
Tables in R, goodness of fit, independence, the expected-count rule, Yates and Fisher, standardized residuals, Cramér's V and McNemar for paired data.
8 lessons · Intermediate
🪜 Nonparametric Tests
Wilcoxon, Mann-Whitney, Kruskal-Wallis, Spearman, the sign test and a hand-built permutation test — what ranks buy you and what they cost.
8 lessons · Intermediate
🪢 Correlation
Covariance, Pearson, Spearman and Kendall, cor.test, matrices with missing values, partial correlation — and the four ways a correlation lies about causation.
8 lessons · Intermediate
📈 Simple Linear Regression
Least squares by hand and with lm(), every line of summary(), intervals for coefficients and predictions, R² and its limits, assumptions and log models.
8 lessons · Advanced
🧮 Multiple Regression
Several predictors at once: confounding, factors and dummies, interactions, centring and polynomials, nested-model tests, NA handling, predict() and reporting.
8 lessons · Advanced
🩺 Regression Diagnostics
Residual patterns, plot(fit), transformations, Breusch-Pagan and VIF by hand, leverage, Cook's distance and autocorrelation — and what to do about each.
8 lessons · Advanced
⚖️ Model Selection
Overfitting by simulation, train/test splits, k-fold cross-validation by hand, AIC and BIC, step() and its dangers, nested F tests, and choosing by purpose.
8 lessons · Advanced
🎯 Logistic Regression
glm(family = binomial), odds ratios, predicted probabilities, thresholds, ROC by hand, deviance tests — and what a coefficient of −4 actually means.
8 lessons · Advanced
🔢 Count Models
Poisson regression with the log link, rate ratios, offsets for exposure, overdispersion, quasi-Poisson, and the negative binomial and zero-inflation ideas.
8 lessons · Advanced
🪺 Mixed-Effects Models
Clustered and repeated-measures data: the ICC, random intercepts, aov with Error(), shrinkage by hand and what lme4 adds.
8 lessons · Advanced
📈 Time Series Basics
ts objects, decomposition, moving averages, differencing, autocorrelation and exponential smoothing, from AirPassengers to your own dated data.
8 lessons · Advanced
📈 Forecasting
Time-ordered splits, naive baselines, Holt-Winters, ARIMA, prediction intervals, error metrics and rolling-origin backtesting in base R.
8 lessons · Advanced
🧭 Principal Component Analysis
prcomp(), scaling, eigenvalues, loadings, scores, the biplot, reconstruction by hand and when PCA fails.
8 lessons · Advanced
🫧 Clustering
Distances, k-means, the elbow and silhouette, hierarchical clustering and dendrograms — and how to tell whether the clusters mean anything.
8 lessons · Advanced
🌳 Trees and Random Forests
Gini and entropy by hand, the best split, growing and pruning a tree, bagging, out-of-bag error and the random forest idea — built in base R before you call rpart or randomForest.
8 lessons · Expert
🎲 Simulation and Bootstrapping
Monte Carlo estimation, bootstrap intervals, permutation tests and power by simulation — inference when there is no formula, and checking the formulas when there is.
8 lessons · Expert
🏁 Capstone Project
One data set, start to finish: question, plan, cleaning, model choice, diagnostics, validation, effect sizes, a clear table and plot, and a reproducible report.
8 lessons · Expert

Start with the free course.

If it teaches you something you can use tomorrow, the rest are waiting.

Open R Basics