Trustworthy Online Controlled Experiments (opens provider site in a new tab)
The A/B testing reference from Microsoft, Google, and LinkedIn veterans. Essential for product analysts.
Cleaning, analyzing, and communicating data with Python or R.
16 picks · 10 free · 13 providers · curated by Rowland Kuru
Editor's intro
Data science is less about fancy models than most course marketing suggests. Day to day, it's getting messy data into shape, asking a sharp question, answering it honestly, and explaining the answer to people who don't read code. This feed weights its picks accordingly: data wrangling and communication get as much space as machine learning.
Pick a language first — Python if you also want to build software or do machine learning, R if you're headed toward research or statistics-heavy work. Learn one data library deeply, then practise on real datasets every week. The certificates here are useful for career changers; the free books are what working data scientists actually reread.
Start here
Python for Data Analysis is written by Wes McKinney, the creator of pandas, and it's free to read online in its third edition. It teaches the tools you'll use every single day — NumPy arrays, pandas DataFrames, grouping, reshaping, time series, and basic plotting — with realistic examples rather than toy ones. Read it alongside a Jupyter notebook and retype the examples on a dataset you care about. It's not a statistics book, so pair it with the Statistics feed once you're comfortable.
The A/B testing reference from Microsoft, Google, and LinkedIn veterans. Essential for product analysts.
Partner Some links on this page are affiliate links, marked Partner. Commissions fund the site at no cost to you. Read our affiliate disclosure.
Loading, cleaning, summarizing, and charting data. Choose Python or R and stick with it for now.
NumPy, pandas, and visualization projects checked by automated tests. Free certification included.
Short seaborn exercises that get you making sensible charts quickly.
A few hours of hands-on exercises. The quickest way to get comfortable with DataFrames.
Hadley Wickham's tidyverse book. If you choose R, this is where to begin.
Spreadsheets, SQL, Tableau, and R with a career focus. Recognized by many entry-level employers.
A Python-first alternative to Google's certificate, with more machine learning toward the end.
Math, statistics, Python, and ML in one package. Breadth over depth, at a modest price.
Full analysis projects, communicating results, and the statistics that keep conclusions honest.
NumPy, pandas, Matplotlib, and scikit-learn in one free, clear, and complete book.
Cole Nussbaumer Knaflic on charts that persuade. The blog and practice exercises are free.
A new real dataset every week. Practise analysis and share results with a friendly community.
Ten R-based courses ending in a capstone. Old-school, rigorous, and still respected.
Rafael Irizarry's R-based program. Strong on inference, modeling, and reproducible analysis.
Causal inference, experimentation, and the data engineering that feeds everything else.
Causal inference for people who need answers, not correlations. Free online and very readable.
A free cohort course on pipelines, warehouses, and orchestration. Where analysis meets infrastructure.
The A/B testing reference from Microsoft, Google, and LinkedIn veterans. Essential for product analysts.