To learn more about graphs, I read ‘The Visual Display of Quantative Information’, a book written by Edward Tufte, an American statistical professor who calls useless charts, ‘chart junk’.
The title of his book ‘The Visual Display of Quantative Information’ does not sound very appealing. the book however, contains beautiful illustrations and is clearly written by someone who is passionate about the topic. I read it from A to Z in no time.
The book made me look at graphs differently. Usually I take graphs for granted. I see them without really looking at them. The main reason is that most of the graphs in newspapers and annual reports don’t look very attractive and are not very useful.
This graph in the 2009 CSR report of Australian bank, CBA, is a good example of a junk graph. It’s a basic time-series bar graph. The graph displays the lost time injury frequency rate over a three years period.
To demonstrate why the graph is not very useful I use Tufte’s method to calculate the data density of the graph. Tufte calculates this as follows:
Number of entries in data matrix / area of data display
The CBA graph presents a data matrix of six entries: three numbers and three years. That’s it. In the report, the graph covers 56 square centimeters, resulting in a data density of only 0.11 numbers per square centimeter.It may be that this graph shows a direction but it also has a very low data density. In Tufte’s words, this is a good example of a ‘junk graph’.
Nowadays anyone can make a graph using with Word, PowerPoint or Excel. In fact, it has become so easy, we all believe we are ‘graph experts’. In Word all you need to do is click on insert chart, click a chart type and select data row. Enter and done! However, this is not the way to make a graph. To do it right you need to know something about statistics and this is what most people (including myself) lack.
Besides good knowledge of statistics, the key to making a good graph is the availability of data, lot’s of it. But I realize that the availability of quantitative data in CSR is still problematic. CSR data collection is not yet high on the agenda of the average CSR department.
To make CSR activities more transparent, companies could perhaps focus more on information gathering. Using more data in the reports and displaying it the right way by using interesting graphs is a logical way forward.
In the next post, I will provide some examples of good and bad graphs in CSR reports. If anyone has a good example of a CSR graph to share, please let me know.