Environmental Economics Article -  Global Warming: Man versus Sun? Environmental Economics website published this global warming article on February 29, 2008.
The paper demonstrates effective data analysis and the use of charts to demonstrate analysis points.
From a data visualization standpoint, the charts can be improved by removing the Excel default background color and heavy grid lines.
This workbook included Tim Haab's data sources and recreates the charts presented in the article, removing the chartjunk, cleaning up axis number formats, and adding calendar values instead of month numbers.
I have posted this workbook so that users can see an example of an effective data oriented investigation and analyze this global warming related data on their own.                   (Workbook link)

I invited Ravens Fan to send me his workbook. He sent it to me. Here's my diagnosis, solution to his problem and Raven's Fans report on how much time he saves each month using dynamic range names.
 

Dynamic range names are critical to avoiding spreadsheet hell and  updating routine charts.

The Ravens Fan spent 1.5 days per month updating his charts before surgery. He reports that his time has been reduced to 10 minutes per month after surgery.

E-mail from Ravens Fan:

"Just wanted to thank you for all your help. I finally figured out the reasoning.  ... What you gave me works great! It took my work load down from 1 1/2 days to 10 minutes. I greatly appreciate your help" ... Ravens Fan

Follow this link to see how dynamic range names can save  you time by automatically updating your charts as new data is entered.

Legends make it harder for the reader to interpret the chart because he/she must move their eye back and forth between the series and the legend. 

Let's replace the legend with data labels for each series to make it easier to interpret. We don't want our message to get lost in the legend.

The labels are added with a simple VBA procedure included in this file. 

This is an interesting question, because WLMPilot wants to draw a line based on a name. The runway numbers drop the last digit of the 3-digit heading, so 07 is actually 70 degrees..

We can use an  XY chart to plot our lines. Since it only takes 2 points to define a line, we can use 0,0 to define one end and calculate the X,Y coordinates for the 2nd point with simple trigonometry functions.  We can then plot our line with the start and end points.  This is the link to revision 2 of the workbook.

The resulting workbook is available here. It provides several features:
1.Drop down menus for the data sheet
2. Drop down menu for variable
3. Scroll bar for minimum X axis value
4. Drop Down menu for X axis time increment
5. Macro to adjust X axis for min value & increment
6. Macro to prepare list of data sheets

BP's chart shows four factors: year, region, and share of world energy use and the equivalent tons of oil equivalent (toe). This graphic has several shortcomings:

• Reader must scan back and forth between pie charts to assess trend in regions share of energy use
• Increase in global energy use from 8.2 to 10.9 billion toe shown directly with text notes and indirectly by adjusting pie chart sizes.
• Important trend information hidden in pie charts:
  • 3 countries have decreasing shares, 3 countries have increasing shares
  • China and Other EMEs have fastest rate of increase in energy share
  • FSU has greatest rate of decrease in energy use
  • US and Europe have decreasing shares of energy use
  • Other OECD countries had slight increase in share of energy use

The Chart Doctor's version is a simple trend chart with each region labeled and the 1991 and 2006 data values displayed. The increasing trend regions are noted with the black line, the decreasing share regions noted with the gray lines.

Initials:
OECD - 30 developed economies, including USA and Europe
EME - Emerging Market Economy countries
FSU - Former Soviet Union countries
Other EMEs - EMEs less China and FSU countries

Jon's chart shows 3 factors in a 2 dimensional (X and Y),  year,. distance from 1 Central Square and % of annual crimes by location. He chose to use a clustered column chart.

Clustered column charts are a poor choice for showing trends of several series because it is more difficult for the viewer to follow the trend pattern for a particular location, the columns for the other series break up the pattern, forcing the reader to work harder than necessary.

The Chart Doctor's Excel version uses a simple XY - scatter trend chart of the 3 series. It clearly shows the abrupt changes in 2006, the decrease in the < 05. mi series and the increase in the > 1 mi series. This version labels the last point so that the legend can be removed and removes the data labels to help focus the viewers attention on what is important, the pattern shift in 2006.

Selecting the right chart type is one of the most important steps in making an effective chart.

Junk_Charts made a "profile chart"  shown in the before column.

The Chart Doctor's Excel version uses a series of XY charts with error bars to produce a panel like display so that readers can compare regional changes by crime type. 

Panel charts allow users to create multivariate plots all in one Excel chart. My is to use a Horizontal Panel Dot Plot. 

JunkChart solved most of the problems, except for information on strength and type of relationship between impervious cover and fecal coliform counts and labeling the creek names.

Here's my Excel version, building on JunkChart.com's initial surgery.

BP does a great service by preparing the annual BP Statistical Review of World Energy and the companion Excel workbook. BP uses stacked/clustered bar charts in their 2006 review. I will be commenting on several of them. In this case, let's look at BP's Oil Consumption Growth for 2003 - 2005 stacked bar chart. Since the bars are cumulative, to follow the growth trends for China and the USA in BP chart, the reader must do some quick mental arithmetic.

My version, uses a simple XY chart  so that each area has a separate series so that the reader can see the US, China and Rest of World values directly without and mental calculation. this version uses a custom Y Axis and includes attaches labels to the individual series rather than use a legend.

The Excel workbook is here if you'd like to create your own version. Be sure to drop a comment if you have suggestions, thoughts or an alternative version.

Let's look at Data360's chart on automobile CO2 emissions. There are a number of concerns:
o Vertical alignment of vehicle labels makes them difficult to read
o Vehicle type (hybrid, SUV) hidden in vehicle name
o Data labels repeat manufacturer
o Emission comparison by manufacturer, not vehicle type
o Horizontal grid not necessary

My Excel chart version uses a dot plot with a 2 level Y axis data label so that the vehicles can be grouped by vehicle type. This version helps to uncover hidden information that was hidden in the Data360 version.