New
Storytelling with Data
A Practical Guide to Communicating Effectively with Data Visualizations and Charts
2025
View →
Pages
317
Published
2017
Practical Statistics for Data Scientists
Statistical Methods and Concepts Every Data Scientist Needs to Know
Build a working command of the statistical methods that actually show up in data science, from sampling and distributions to regression and machine learning.
Most statistics textbooks were written for mathematicians, not practitioners. This book flips that. Peter Bruce and Andrew Bruce cut through academic formality to focus on the concepts and techniques that matter in real data science work. Covering probability, sampling, statistical experiments, regression, and classification, it gives you the mental models to reason clearly about data — and the vocabulary to communicate what you find.
About this book
Statistics is the foundation of data science, but most introductions to the field bury the signal in academic notation and proof-heavy explanations. This book takes a different approach. It focuses on the concepts that actually appear in practice — the ones you need to understand before you can reason correctly about data, build models, or interpret results.
Peter Bruce and Andrew Bruce have worked in applied statistics and data mining for decades. The result is a book written from the practitioner's perspective: what you need to know, why it matters, and how to use it. The examples are grounded in real data. The explanations skip the unnecessary formalism without dumbing anything down.
You will start with exploratory data analysis — summary statistics, distributions, and visualizations that tell you what your data actually contains before you start modeling. From there the book covers the mechanics of sampling and bias, the logic of statistical experiments and A/B tests, and the probability theory you need without the graduate-course overhead. The second half moves into regression, classification, and the statistical ideas that sit underneath common machine learning methods.
Each topic is organized around a clear definition, key terms, and worked examples in both R and Python. The structure lets you read cover to cover or jump directly to the concept you need right now.
- Exploratory data analysis: location, variability, distributions, and correlation
- Sampling, bias, and the selection effects that break real analyses
- Statistical experiments, A/B testing, and hypothesis testing done correctly
- Probability distributions and their practical uses
- Regression — linear, multiple, and the assumptions you must check
- Classification methods including logistic regression, decision trees, and Naive Bayes
- Resampling, bootstrapping, and cross-validation as practical tools
If you are a developer or analyst who works with data and knows you need a stronger statistical foundation, this is where to start. It is rigorous enough to be useful and practical enough to read in a weekend.
🎯 What you'll learn
- Summarize and visualize datasets using the right measures of location, variability, and distribution shape
- Identify and account for bias in sampling before it corrupts your analysis
- Design and interpret A/B tests and hypothesis tests without common logical errors
- Apply probability distributions to model real-world uncertainty in data
- Build and diagnose linear and multiple regression models, including checking the assumptions that matter
- Use classification algorithms — logistic regression, decision trees, Naive Bayes — with a clear understanding of their statistical basis
- Apply bootstrapping and cross-validation to get honest performance estimates from limited data
- Translate statistical findings into plain language conclusions that hold up under scrutiny
👤 Who is this book for?
- Software engineers transitioning into data science who need statistical grounding without a formal academic detour
- Data analysts who apply statistical methods daily but want to understand what is actually happening under the hood
- Machine learning practitioners who can tune a model but want clearer intuition about bias, variance, and significance
- Business intelligence professionals who run A/B tests and need to interpret results with more confidence
- Python or R programmers who have avoided statistics and want a practitioner-focused entry point
Table of contents
-
01
Exploratory Data Analysis
You learn how to summarize and describe datasets using measures of location and variability, and how to spot patterns, outliers, and distribution shapes before any modeling begins.
-
02
Data and Sampling Distributions
You work through the principles of random sampling, selection bias, and the central limit theorem, building the foundation for every inferential technique that follows.
-
03
Statistical Experiments and Significance Testing
You learn the mechanics of A/B testing, hypothesis testing, p-values, and confidence intervals, with clear explanations of what each of these actually tells you — and what it does not.
-
04
Regression and Prediction
You build linear and multiple regression models, learn to interpret coefficients correctly, and check the residual diagnostics that distinguish a trustworthy model from a misleading one.
-
05
Classification
You apply logistic regression, linear discriminant analysis, Naive Bayes, and decision trees to classification problems, and learn how to evaluate classifier performance beyond raw accuracy.
-
06
Statistical Machine Learning
You explore tree-based ensemble methods, bagging, and random forests, connecting modern machine learning tools back to the statistical principles that explain when and why they work.
-
07
Unsupervised Learning
You apply clustering methods including K-means and hierarchical clustering, and use principal components analysis to reduce dimensionality in high-dimensional datasets.
Frequently asked questions
Do I need a statistics background to read this book?
No. The book assumes you are comfortable with basic algebra and have some experience working with data, but it does not require prior coursework in statistics or probability.
Which programming language does the book use?
Examples are provided in both R and Python. You do not need to be fluent in both — following along in one language is enough to get full value from the worked examples.
Is this book still relevant given it was published in 2017?
The core statistical concepts covered — sampling, regression, hypothesis testing, classification — are not framework-dependent and remain fully valid. Specific library syntax may have evolved, but the reasoning and methods translate directly to current practice.
Is this a theory-heavy or a practical book?
It leans firmly practical. Formal proofs are avoided in favor of clear definitions, worked examples, and intuition-building explanations aimed at practitioners rather than statisticians.
Who is this book not for?
It is not the right starting point if you want rigorous mathematical statistics with proofs and derivations. For that, a graduate-level textbook would be more appropriate.
You might also like
New
New
Data Science: The Hard Parts
Techniques for Thinking Analytically and Solving Real Data Problems
2023
View →
New
Fundamentals of Data Engineering
A practical guide to the complete data engineering lifecycle, from ingestion to serving
by Joe Reis, Matt Housley
2022
View →
New
Data Analysis with Python and PySpark
A hands-on guide to scalable data analytics using Python and PySpark
2022
View →