Understanding Data: A 21st Century Approach to Statistics and Data Science

What Makes This Statistics Book Different?

Most statistics textbooks still ask students to memorize formulas built for an era when all calculations had to be done by hand. Alan Garfinkel and Yina Guo take a fundamentally different path. Understanding Data uses simulation and resampling as the core learning engine, letting students see statistical concepts in action rather than simply applying recipes. The result is a deeper, more intuitive grasp of probability, inference, and data analysis that stays with readers long after the course ends.

Starting from the Reproducibility Crisis

The book begins where modern statistics must: with the crisis of unreliable and irreproducible results in the biomedical and social sciences. Garfinkel and Guo explain why the concept of statistical significance has been so damaging and why p-hacking and formula worship have contributed to the public loss of trust in science. Instead of quietly teaching the old methods, this text equips students to understand the problems and adopt better practices from the start.

What the Book Covers

• Describing and presenting data using dot plots, beeswarm plots, histograms, kernel density estimates, and more
• Modeling and simulating probabilities to build real intuition
• Null hypothesis significance testing—including a clear-eyed discussion of its flaws
• Confidence intervals through resampling, with no reliance on harsh mathematical assumptions
• The normal distribution and why real-world data often departs from it
• Comparing two groups using effect sizes, ranks, and paired study designs

Each chapter moves from concepts to computational practice, always showing the data and the reasoning behind the numbers.

Who This Book Is For

Written for first-year undergraduates in the biological, biomedical, and social sciences, Understanding Data requires no mathematical background. It is equally suited to introductory data science courses and to anyone who wants a modern, simulation-first understanding of statistics. Readers will come away with the ability to think clearly about uncertainty, visualize data honestly, and avoid the statistical traps that have undermined so much published research.

A Better Way to Learn Statistics

By replacing hand-calculation recipes with computational resampling, this book opens the door to asking more relevant scientific questions. You are no longer restricted to the narrow cases where formulas happen to work. Instead, you gain a flexible, widely applicable method for drawing meaningful conclusions from data—exactly the skill today’s data-driven world demands.