# Effective Graphs with R

September 13, 2012
By

(This article was first published on Revolutions, and kindly contributed to R-bloggers)

Today's guest post is by Naomi Robbins, author of the Effective Graphs blog  ed.

I write a blog on effective graphs for Forbes. David Smith
invited me to write a guest post here that was a roundup of some of my Forbes
posts where R was used. My use of R graphics has ranged from simple box plots
and dot plots to examples of diverging stacked bar charts drawn with the likert
function of the HH package.
The audience for this blog is the general public rather than statisticians and
scientists although many of the latter regularly visit.

In Gelman's
Click-through Compromise for Statistical Graphics and Data Art
I show a
superposed dot plot as a statistical graphics alternative to an artistic
visualization of revenues and profits of the Fortune 500. At a panel entitled “What Makes
Good Data Visualization,”
the panelists stated that artists and
statisticians have different goals in creating visualizations. Andrew Gelman
said that artists want visualizations to be creative, grabby and innovative
while these characteristics are not the goal of statisticians. Gelman suggested
a compromise: a click-through future where we can have a figure that grabs your
attention and gets you interested. You click through to a statistical graph
that is clear, accurate, and easy to understand. Click through again and see a
spreadsheet with the data. The superposed dot plot shown in the post (and below) uses code
similar to that of Cleveland’s barley example

One of my pet peeves is seeing spaces between the bars of a
histogram or seeing histograms called bar charts. In A
Histogram is NOT a Bar Chart
, I point out
the differences between bar charts and histograms. The code used for these
figures is a simple use of hist. That post was followed by Comparing
Distributions with Box Plots
, which shows the superiority of box plots (such as shown below) over
histograms for comparing distributions. The code was straightforward uses of
the lattice package. When
Should I Use Logarithmic Scales in My Charts and Graphs
is another post
that makes straightforward use of dot plots with logarithmic scales.

The remaining two posts that I’ll discuss here include
diverging stacked bar charts drawn using the likert function of the HH package.
Alternative
to Charles Blow's Figure in “Newt's War on Poor Children”
, written jointly
with Rich Heiberger, supports an argument in an op-ed article in the New York
Times written by Charles Blow. We redesigned the figures used by Blow to make
what we consider to be a stronger presentation with a simpler graphical style.

Thinking
discusses how chart menus in software discourage
thinking about the data and often cause the message the graph designer intends
to convey to be lost. In one of the examples I show a before figure (below) that used six pie charts to compare the age
distributions of users of six brands of a product. My after figure was a diverging stacked bar chart. This post is my favorite
one because it questions standard advice and encourages readers to think rather

Naomi Robbins is a consultant and seminar leader who specializes in the graphical display of data. She trains employees of corporations and organizations on effective data visualization. Naomi also reviews documents and presentations for clients, suggesting improvements or alternative presentations as appropriate. She is the author of Creating More Effective Graphs, published by John Wiley (2005). In addition to her one and two day seminars on creating more effective graphs, Naomi offers short programs such as “Recognizing Misleading and Deceptive Graphs” and “How to Avoid Common Graphical Mistakes.” Naomi received a Ph.D. in mathematical statistics from Columbia University, M.A. from Cornell University, and A.B. from Bryn Mawr College. She had a long career at Bell Laboratories before forming NBR, her consulting practice.

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: Data science, Big Data, R jobs, visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...

Tags: , ,