Climate Central - A Failure in Graphics and/or Statistics

Recently, Climate Central issued a new report that discussed warming across the continental US over the past 100 years.  In this report, the graphics and commentary are extremely misleading.  For example, even though 1/2 of the state's climate trend did not meet statistical relevancy, the created every single chart as if it did.

In their original graph - note the "warming rates accelerated everywhere" title.  This is cherry picking of the data.  In fact, their statistically insignificant data showed that Alabama, Arkansas, and Georgia COOLED.  (Not that I believe it).  In one case, the "warming" was by 0.008 degrees.  Whew...  I'm fixin' to pass out!

Anyway, here is my revised graph using only the data that was statistically relevant (see footnote in table 3 of the report that you can download [Heat Is On Report].  If you're going to make a point, you have to use honest graphics.  So, yes, there is some real warming... but not in 1/2 of the states.

Next, they tried to show which states were warming the fastest.  Of course, the title of the chart uses the terms "10 Fastest Warming States" and "10 Slowest Warming States".    I corrected that for them, too.

Take a look... Last I heard, when something was declining in temperature, you don't call it warming.  There's another word for that.  Cooling?  Of course, since the data was statistically insignificant, why even create this graphic?

I'm a huge fan of Edward Tufte.  Visual Display of Quantitative Information  Whomever created the graphics for the Climate Central report really should find some of his work and use it to improve the way they display data.  As Tufte said in a seminar I attended, "Clear drawings are clear thinking..."  We could all use more of that.