This post may contain affiliate links. Please read my disclosure for more info:

How to create misleading charts

hot to create misleading charts

hot to create misleading charts

Visualization experts such as Stephen Few and Stephanie Evergreen inspire us to make charts which enable users to understand information quickly and accurately.  Personally, I love articles about how to make charts more usable.  But there is also the other side, the marketing activities, the political organizations and maybe your boss, who wants to put a bit of ‘spin’ into the charts.  Is it right to purposefully create visually misleading charts?  Well . . . that’s for you to decide.  By looking at how to purposefully create visually misleading charts we can (a) understand the tactics which others might use (b) learn to avoid the bad practices which might cause charts to be misleading (c) can make visually misleading charts, if we are ever called upon to do so.

Example 1:

Look at the chart below – that company is certainly experiencing some fantastic growth!  Or so you are meant to believe.

Visually Misleading Charts - example 1

However, this purely the trick of perspective.  The bars at the front look much bigger than the bars at the back.  Also, the perspective has been forced to make the bottom of the charts appear level, this makes the bars at the front appear even bigger.

The use of color in this chart is also intended to deceive.  The color green is usually associated with good results; it tells you that all is on track.  By including an upwards arrow turning from amber to green this is informing you that results are on the rise.  As the amber and green is more vivid that the grey bars your eyes are drawn to the arrow, rather than the bars.

If we were to view this as a simple flat column chart we would see this:

Visually Misleading Charts - example 1
This chart reveals a company which had seen small growth for 2 years, but then has seen a small decline.  Actually, 2016 has the same result as 2012.  Look back at the first chart, does the 2016 result look even close to 2012?  No, it doesn’t.  It was certainly a misleading chart.

 

Example 2:
Wow! Company A looks about twice the size of Company B, and at least four times bigger than Company C.

The key to this chart is that the axis does not start at zero.  The purpose of a bar or column chart is to show to relative sizes of each bar.  However, by setting the base at 25, rather than zero it makes Company A appear significantly bigger.

Even though the axis has been labelled, it is show in a very light shade, so as not to draw attention to itself.  Also, the bottom axis line is thicker than the other gridlines, this gives the impression that this has a zero base.

If we were to view this as a simple flat column chart we would see this:

Visually Misleading Charts - example 2

Yes, Company A is the biggest, but not by the magnitudes shown in the chart above.

 

Example 3:
Clearly, Product A is much more popular that Product B and Product C – it must be, as it’s slice is so much bigger.

Visually Misleading Charts - example 3

So, what are the tricks at play here?

Firstly, by using a 3D effect, Product A occupies a greater proportion of the screen and therefore appears bigger.  By creating a 3D effect the bottom section always occupies a greater proportion due to the side being visible.  This gives the illusion of greater area.  Product A has been purposefully positioned at the bottom of the chart to achieve this effect.

The colors used also make Product A appear more vivid, your eyes are not even looking at Product B or Product C.

Finally, the data is presented as a pie chart.  Whilst a very common visualization tool, it is very difficult to read and understand correctly.  Where the lines are at 0, 90, 180 or 270 degrees it is very easy to estimate the size of a section, but once rotated, it becomes much harder to mentally calculate the real values.

If we were to view this as a simple flat column chart we would see this:

Visually Misleading Charts - example 3

Wow! Look at this.  Product A has gone from appearing to be the biggest slice on the pie chart to being the smallest column.  We have not changed the values; we have simply applied a suitable visual presentation.

 

Example 4:
Look at the following, do you have any idea how much better Salesman A is doing the Salesman B?

Visually Misleading Charts - example 4

Salesman A is certainly better than Salesman B and Salesman C, but that’s about as much as we can tell.  They key here is that it is relying on area to represent the values.  The human brain is very bad at comparing area, so it would purely be a guess as to what the relative sizes were.

If we were to view this as a simple flat column chart we would see this:

Visually Misleading Charts - example 4

Only once we can see the height of the bars (which the human brain is good at comparing) is it’s possible to see how much better Salesman A is than the others.

 

Conclusion:

If you ever wish to make visually misleading charts the keys to success are:

  • Use perspective to make the items closer to the front bigger
  • Use positive or negative colors, even if there is nothing positive or negative being shown
  • Use an axis which starts at a number other than zero, and do not emphasize that it starts at a non-zero value.
  • Use 3D effects to make close items appear bigger
  • Use Pie Charts and make none of the lines are at 0, 90, 180 or 270 degrees
  • Use areas to show values, as nobody will be able to process what this really means

 

What Next?

Do you want to create beautiful, easy to read and insightful charts all the time?  If so, I high recommend you get Effective Data Visualization: The Right Chart for the Right Data(affiliate link).  This is one of my favorite Excel books.

Not sure if this book is for you?  Read my review.

Save

Leave a Reply

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