Погрешность: статистика для мира, которому нужны лучшие решения

Statistics is how we figure out whether a treatment works, whether a policy is helping or hurting, and whether the pattern we see in data is real or a mirage. Margin of Error is a modern introduction to statistical thinking — rigorous enough for graduate use, accessible enough for students who walked in convinced they were "not a math person."Built around real stories with real consequences — the Flint water crisis, résumé-discrimination audits, election polling, COVID testing, and the gender wage gap — this book teaches the reasoning behind the methods, not just the formulas. Every chapter opens with a story, develops the statistical idea it raises, and closes by asking whether the analysis was done responsibly.What you'll learn: How to summarize, visualize, and reason about data How to summarize, visualize, and reason about data Probability, the normal distribution, and the central limit theorem — explained in plain English Probability, the normal distribution, and the central limit theorem — explained in plain English Confidence intervals and hypothesis testing without the cookbook feel Confidence intervals and hypothesis testing without the cookbook feel Comparing groups, simple and multiple regression, and how regression actually gets misused in practice Comparing groups, simple and multiple regression, and how regression actually gets misused in practice Where AI tools get statistical reasoning wrong, and how to spot it Where AI tools get statistical reasoning wrong, and how to spot it Designed for business students, working analysts, and anyone who needs to evaluate quantitative claims in the wild.Companion materials — interactive applications, downloadable datasets, and a chapter-by-chapter R Companion that reproduces every analysis — are free at stats.marginoferrormedia.com.By Vivek H. Patil, Ph.D. — Professor of Marketing at Gonzaga University and author of 25+ peer-reviewed articles journals including the Journal of Business Research, PLOS ONE, and the Journal of Marketing Analytics.A Margin of Error Media publication.