(This article was originally published at Effective Graphs, and syndicated at StatsBlogs.)
What an outstanding group of readers I have! Both the quality and the quantity of entries to the Graph Makeover Contest were amazing. There were 29 entries to the first part of the contest which consisted of presenting alternatives to a pie/donut chart and 28 entries to the second part of the contest which was visualizing a table. Today we will present some of pie/donut alternatives and my next post will discuss visualizing the table.
Figure 1 shows the original chart.
Figure 1. Source: http://demand.eloqua.com/40-must-see-charts
Readers criticized a number of features of the original figure in addition to the fact that donut charts do not communicate as clearly as some other graph forms. Several pointed out that the statement “LinkedIn referral traffic is 16x higher for B2B companies” is not necessarily true since we don’t know the sample sizes. One percent of a very large number can be larger than 16% of a much smaller number. It is true that “The proportion of LinkedIn referral traffic is 16x higher for B2B companies.” Others mentioned that the data emphasize the dominance of Facebook. They question why emphasize the LinkedIn comparison when LinkedIn is a business service. I would expect the proportion of LinkedIn traffic to be at least that much higher for B2B companies. What surprised me was that Facebook traffic was as large as it was for B2B companies.
Of the 29 entries, 20 were variations of bar charts. There were both horizontal and vertical grouped bar charts, some grouped by B2B and B2C and some grouped by the social media source. There were both horizontal and vertical divided bar charts, a back-to-back bar chart and a diverging stacked bar chart. There even was a dynamic bar chart showing the streaming nature of the data. In addition, there were three slope charts, three dot plots including both horizontal and vertical ones, a nested donut chart, a sentence and a waffle chart (also called a squared pie chart.)
Please comment on the article here: Effective Graphs
