Graphics to explain data need to be both effective and attractive: introducing icon graphs

A recent article made two points regarding the graphical presentation of information. The points derive from a study that presented four kinds of graphics and asked a group of doctors to rate a pair their characteristics: were the graphs correct and were they enjoyable to look at.  The doctors said that an “icon graph” was the most accurate depiction of the data, compared to a table, a bar graph, or a pie chart, though it was also the least enjoyable.


Since icon graphs are new to me, let me describe the one the article showed.  The icon graph had for rows and four columns of squares.  Some of the squares were solid green, some solid red, and some had a diagonal line through them from the top left to the bottom right. Thus, each cell conveyed one of four combinations: either color and lined or not.  That’s a lot of information to convey in one graphic!


For example, if you had four categories of law firms you retained last year, by number of lawyers, and you had several ranges of fees paid, an icon graph could locate the appropriately colored/lined square for each firm over its fees paid during the year.


You could add even more complexity (such as three colors or bold outline dotted and solid diagonals) and thereby show many more variables about the law firms, but the risk would be that those who have to make sense out of the icon graph would fail.



Take the GC Metrics 2013 benchmark survey, all you graphics fans, and enjoy the various graphics in Release 1.0  Here is the URL:  It’s free, quick, and you will get five Releases.  The survey asks for six of your 2012 figures: number of lawyers, paralegals, and other staff; inside and external legal spend; and revenue.

We welcome comments

Your email address will not be published. Required fields are marked *