Survey Software, Web Survey, Online Surveys, and Enterprise Feedback Management solutions from Vovici

Your email:
   

Welcome to the Listening Post!

Your single source for everything Voice of the Customer (VoC) and Customer Experience (CxP). And, don’t forget you can follow us on twitter @vovici, or come check us out on Facebook and join the Vovici Network on LinkedIn.

 

Current Articles | RSS Feed RSS Feed

Pie Charts Have No Place in Market Research

 

The goal of conducting a survey or other market-research study is to come up with useful information and communicate it effectively. Pie charts miscommunicate data and therefore should be avoided.

From the following chart, can you list the products by order of preference for each question?

pie charts bad

You can’t, and it’s not your fault. Our visual perception does not always accurately reflect what we are seeing: we have difficulty comparing and contrasting visual area, we struggle when comparing angles, and with pies we are forced to compare wedges that aren’t adjacent to one another:

  • Stanley Smith Stevens studied the limits of human perceptions across many different dimensions. He found that we accurately perceived differences in brightness, length and cold, but we overestimate small differences in visual area and progressively underestimate large differences in area. (See Stevens’ power law for more details.)
  • William Cleveland, in The Elements of Graphing Data, relates the results of an experiment where 51 subjects judged 40 pairs of values on bar charts and then on pie charts. In almost all cases, the data was interpreted more accurately using bar charts (the exceptions were pie charts that compared one slice to the whole when that slice was 25%, 50% or 75%).
  • As Edward Tufte writes in The Visual Display of Quantitative Information, “A table is nearly always better than a dumb pie chart; the only thing worse than a pie chart is several of them, for then the viewer is asked to compare quantities located in spatial disarray both within and between pies. Given their low data-density and failure to order numbers along a visual dimension, pie charts should never be used.”

Here’s a comparison of the same data shown with pie charts and column charts:

pie charts vs bar charts 2

The bar charts ("column charts" in many graphing packages) make the small differences in products easily recognizable. Pie charts needlessly confuse.

For far more entertaining takedowns of the pie chart, see Annie Pettit’s “Pie Charts – Our Evil Friend” and Coda Hale’s Google Analytics rant, wherein he says, “Pie charts are the information visualization equivalent of a roofing hammer to the frontal lobe. They have no place in the world of grownups, and occupy the same semiotic space as short pants, a runny nose, and chocolate smeared on one’s face.”

Chocolate, or pie, perhaps.

Comments

Absolutely right, I couldn't agree more! I see pie charts in research reports far too often...
Posted @ Wednesday, March 30, 2011 10:19 AM by Ruth Stevenson
When I saw the title of this blog in my inbox, I thought 'What? Why ever not?!' and had to read super quickly because I was in the middle of creating a pie-chart, as it goes. 
 
Now I will create a bar graph instead! 
 
However, from a 'how engaged are the Cients' point of view, I reckon they'd get pretty bored of seeing bar graph after bar graph, and from one report i've seen recently, it looks like some people put in any old graph for the sake of a bit of variance and not taking into consideration how 'readable' the graph is, or even what on earth it's trying to communicate. 
 
Maybe graphs themselves are the problem? How else can we present findings, (i.e - the meaty stuff that matters, the information that Clients can actually DO something with) and NOT data without baffling the Clients? 
 
Perhaps the future is in video reports... 
 
Betty
Posted @ Wednesday, March 30, 2011 10:22 AM by Betty Adamou
Jeff, Hi. 
 
Love the Voice of Vovici normally. In this case, sorry, I disagree. Not to disparage bar or line graphs or the many ways data can be shown, but sometimes nothing beats the simplicity of a pie chart. And, uh, re: the Coda Hale blog: "anything with a piechart reeks of bozocity" coupled with "we're trying to act like grownups"? Really? 
 
Also, the bar charts might be easier to interpret because you've shown percentages.  
 
Best regards, 
 
Chris
Posted @ Wednesday, March 30, 2011 11:00 AM by Chris Svoboda
Here, here. This is one of my pet peeves as well. I almost never use them.
Posted @ Wednesday, March 30, 2011 2:55 PM by Gene Saravu
Pie charts work much better than bar charts when there are about 2 to 4 items/slices and you want to communicate that the numbers add to 100% (e.g., Yes, No, Maybe). A key thing is to put the label and % right on or near the slice.
Posted @ Wednesday, March 30, 2011 4:00 PM by Gary
I hope the real point of the blog is to remind all in the field to pay more attention to aligning the intended and visual interpretations rather than convince us to totally ban the use of certain graphic displays. I'd hope that the research community would also avoid showing bar charts which exaggerate what appears to be a 5 percentage point difference on the bar charts. Since there are no labels on the bar charts you could argue that the bar and pie charts are equally misleading in this case, or conclude that at least the bar chart better displayed the finding of “no difference.”
Posted @ Wednesday, March 30, 2011 5:06 PM by Susan Ford
Coda Hale's comment is over the top, but I agree with article about not using pie charts to communicate data. I prefer the horizontal bar charts and use those the most in reports.
Posted @ Thursday, March 31, 2011 1:26 PM by Ellen
A number of people have commented here and elsewhere about the original draft's use of a graphic with an axis showing percents for each bar chart (click on the second graph to see the original). I think even without a labeled axis the bar chart is easier to read (see above), for the reasons given by Stevens, Cleveland and Tufte. 
 
Yes, my commentary is over the top -- that's why I'm a blogger and not an academic researcher. Will our software continue to support pie charts? Of course. Will I use pie charts in research reports? Hopefully not. I plan to restrain my use to thumbnail images consisting of just two slices. 
 
Thanks for all the great comments!
Posted @ Thursday, March 31, 2011 4:32 PM by Jeffrey Henning
Post Comment
Name
 *
Email
 *
Website (optional)
Comment
 *

Allowed tags: <a> link, <b> bold, <i> italics

Latest Posts

Loading
What's New
Don't Be in the 4%
VoC on Twitter
Verint Blog
Verint Blog: Read the Latest from the Verint Systems Blog