« Flaming out | Main | Why were they laughing? »


Naomi B. Robbins

Nice post. I'd call the redone plot a dot plot rather than a bar chart.


i always look at your redone chart first but to be honest, i don't get your chart this time.
i don't get the x-axis, what does it show? how many people who are highly concerned are also concerned?
furthermore i can't see how this chart is better than a table. because of the high white space on the x-axis (roughly an 45% range is unused; and that does not include the fact 0 is not to the very left of the axis), the differences between the categories are low/hard to make out. why bother at all then?

after reading the post, i understand what the x-axis shows. still hard to make out all the differences.


It's not counterfeit, because whoever made the chart probably doesn't realize what they did wrong. It's 'cargo cult' graphics; they don't really know how to deal with numbers, they just have a vague idea of what the end result kind of looks like.


Naomi: Yes, I switched to a dot plot and forgot to change the text. Thanks for pointing it out.

Thomas: There are a few details on my chart that can leave readers a little perplexed if you didn't read the original first. I didn't put the survey question as the title. Also, I made this plot in R which likes to express percentages as a decimal between 0 and 1.
As for "proportion concerned/highly concerned", the original designer is following a well-established tradition in survey research analysis, where we focus on the "top two boxes" of a five-point scale. One can certainly take issue with this but it's an industry standard.

Mitch: the source of the chart seems to be Forrester Research, which should have done better. They certainly are no novice.

Beth Renneisen

Thanks for this site, Len! I'll use the bubble chart in my class, and the messed up bar chart, too. Beth


Ultimate non-self-sufficient chart? It seems that many people compete for the Guinness World Record.

I played with my image analysis tool, keeping only the bubbles (even the head I know) and calculating the objects area (values are scaled based upon the 72% bubble). Results are, as you say, very far from what was written (Cf. http://stephane.vellay.free.fr/images/JunkChartBubblesArea.jpg )

I actually have another idea about the value they tried to follow for the bubble size: the number of letters in the bubble :)

PS: To continue with Naomi's comment, the post is not tagged as dot plot but as bar chart

The comments to this entry are closed.

Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR.

See my Youtube and Flickr.

Book Blog

Link to junkcharts

Graphics design by Amanda Lee

The Read

Keep in Touch

follow me on Twitter