This directory holds files that contain new content, from the draft
for a new text, derivative from "Data Analysis and Graphics Using R",
that omits most of the R tutorial content of the 3rd edition, and
that compresses the substance of the 3rd edition chapters 2 -- 5
content into a single chapter.
A provisional title is:
Data Analysis With R -- Learning by Example.
* changesIN4th.html: Overview of changes, intended for the new
preface.
* draft-toc.pdf: Table of contents
* chap1.pdf: This chapter is designed as a scene-setter for the text
as it now is. It has broadbrush notes on the changing statistical
applications scene, on the role of statistics in scientific
endeavour, on graphics, on data summary, on distributions, on
project management, and on reproducible reporting.
* Ch2-pvalCritique.pdf: Misuse and misunderstanding of p-values is
widespread. The account now given makes a strong point that
p-values are not an effective substitute for the independent
replication of scientific results.
* Ch2-FalseDiscoveryRate.pdf: In contexts where there can be many
p-values (tens, or hundreds, or thousands, or more), new and
productive possibilities emerge for the conclusions that can be
drawn.
* Ch4-QuantileReg.pdf: Models for the quantiles of the response
(e.g., 25% or lower quartile, median, 75% or upper quartile) are
an important and useful extension of the linear model framework.
This new material illustrates the new and exciting possibilities
offered by the abilities of the quantreg package (Koenker, 2015).
* ch4-hurricanes.pdf: An analysis of data on deaths from US Atlantic
hurricanes.
* ch5-negBinErrors.pdf: Example of use of negative binomial family
error.
* ch7-BetaBinomial.pdf: Example of use of beta binomial family
error.
* ch9-RNAseq-FDR.pdf: Analysis of RNA-Seq data -- a nice example
of false discovery rate (FDR) estimation.