Reasoning with Data: An Introduction to Traditional and Bayesian Statistics Using R

by Jeffrey M. Stanton

Engaging and accessible, this book teaches readers how to use inferential statistical thinking to check their assumptions, assess evidence about their beliefs, and avoid overinterpreting results that may look more promising than they really are. It provides step-by-step guidance for using both classical (frequentist) and Bayesian approaches to inference. Statistical techniques covered side by side from both frequentist and Bayesian approaches include hypothesis testing, replication, analysis of variance, calculation of effect sizes, regression, time series analysis, and more. Students also get a complete introduction to the open-source R programming language and its key packages. Throughout the text, simple commands in R demonstrate essential data analysis skills using real-data examples. The companion website provides annotated R code for the book's examples, in-class exercises, supplemental reading lists, and links to online videos, interactive materials, and other resources.
 
Pedagogical Features
*Playful, conversational style and gradual approach; suitable for students without strong math backgrounds.
*End-of-chapter exercises based on real data supplied in the free R package.
*Technical explanation and equation/output boxes.
*Appendices on how to install R and work with the sample datasets. 

325 pages, Hardcover

Published May 19, 2017

Reviews

[Peter Baumgartner · Rating 5 out of 5]

A very gently introduction into frequentist and Bayesian statistics:

1.) The author starts each chapter with fake data so that you know how data were generated and what results you should therefore expect. This is a good educational device to understand the applied procedures better.

2.) The author then applies frequentist procedures to these fake data and explains the result step by step.

3. Subsequently, he changes to real data (taken from internal R data sets) and replicates the procedure. Using internal R datasets has the advantage that you don't have to worry about uploading and data cleaning. The downside — that you do not learn about data wrangling — is known by the author, but was not in his scope for this introductory book.

4.) Finally, he applies Bayesian statistics to these data and explains the differences. The book discusses in very details the resulting R summary outputs, and what the many diagnostic measures mean.

Throughout the book, you learn about numerous R packages supplied by the active community. A small disadvantage is that the book’s code is written in base R and not with the tidyverse approach. The code cannot be copied from the book, but there exists an accompanying website, where you can download all the code snippets.

Despite these small disadvantages, I rated the book with 5 stars because it gave me not only an understandable overview of the most important statistical frequentist procedures but demonstrates convincingly the advantages of Bayesian procedures with the same data.

[Ward · Rating 3 out of 5]

A solid three stars on the textbook rating scale this book is a good overview of introductory statistics and their applications using R. Topics span from the most basic; mean and median to more complex such as time series analysis and principal components. Stanton also explains Bayesian approaches to most of the techniques articulating the main differences between the two.