DeepRead
Cover of Statistical Inference as Severe Testing

Statistical Inference as Severe Testing

Deborah G. Mayo · 2018

Machine learning, statistics and probability › Introductory statistics

#9,498 most recommended

About the book

The book covers the philosophy and practice of statistical inference, focusing on how significance tests, p-values, confidence intervals and related methods are used to learn from data. It addresses the disputes among frequentist, Bayesian and likelihoodist schools, and the replication problems that followed the so-called statistics wars in the social and biomedical sciences. Mayo argues that a claim is supported by data only when the data have passed a severe test, meaning the test would likely have revealed the claim to be false if it were false. She develops this error-statistical account as a way to avoid common fallacies such as treating a nonsignificant result as evidence of no effect or equating a small p-value with a large effect. The book defends pre-data error probabilities against Bayesian and likelihoodist alternatives, and proposes practical norms for interpreting tests, including attention to selective reporting, multiple testing and optional stopping.

Written by Claude from what it knows of this book.

What Hacker News says 67% recommend · 33% critical

01220192020202120222019 peak · 2
This bookAverage book in Introductory statistics

In the commenters’ own words

What Open Library readers say

No Open Library reader has rated it yet.

  • 1 want to read
  • 0 reading
  • 0 have read

Open Library’s page for the book

What critics say, via Wikipedia

No Wikipedia article about this book yet, so no critics to quote.

Reddit communities — coming later

Details

First published
2018
ISBN
9781107054134
Editions known
8 at Open Library
Open Library
OL20172506W

Error analysis (mathematics)Fallacies (logic)InferenceMathematical statistics

Explore in the mosaic

Read next: its neighbours on the mosaic