Y-axis Shortening Fail - Data Viz Renovations
Bad data viz submissions ► https://goo.gl/forms/Y66ByqEQ97
Source ► / so_dutch_people_are_about_3_times_as_big_as
Today we're going to be renovating this infamously bad data viz; focussing on the average height of males in a few countries... At first glance it doesn't look too bad, right? But the more you look at it, the more data viz faux pas you begin to notice.
First off is the Truncated Y axis for bar charts. Whilst there are some exceptions to do this for line charts, it is never ok for bar charts. It masks the true height of the data points and makes the relative difference between them seem larger than they are.
Here’s a tip Tip - if your data points "touch" the bottom of the Y axis, don't truncate it. This is perfectly demonstrated here;
In real life, the Philippines data point's feet wouldn't start at 1.5 meters…, yet here in this data viz they do. It confuses the audience as it makes the Philippines data point visually seem three and a half times smaller than the Netherlands, whereas it's only 12% when using the metrics.
Another issue is with the stereotypes representing each country's data point.
The Netherlands wear clogs
Americans are all fat and dress like skaters
Canadian's are all lumberjacks
All English people wear suits and bowler hats
All Indians were Dhotis and Dastars
And Filipinos… wear yellow t-shirts…?
Actually, they are only normal ones on the data viz
Apart from being 3 times smaller than they are…
The visual stereotypes are bad for 2 reasons;
Firstly, all of the different shapes and colours interfere with the audience's ability to read the data properly. Secondly; stereotypes are harmful and not true in the slightest…
The blue background is not helping the viz in any way either - it's just extra visual clutter getting in the way.
Lastly the title isn't very helpful. Who's looking down on whom? Sure the taller people might have to physically have to look down at the shorter ones, but it could also imply that being taller makes you superior, what clearly isn't the case…
So how do we give this viz a renovation?
First off let's fix an easy item by removing the coloured background. It wasn't helping anyone.
Now's let's remove the visual stereotypes. We can still keep a bit of visual flair, by keeping the male icons, applying a consistent style and having their heads represent their countries' flags. It's now visually consistent with a visual pattern the audience can learn and follow quickly.
Sure we could turn these purely into bars, but this is Data Viz renovations, not Data Viz Minimalism. Plus I'm assuming the target audience for this data viz is pretty casual, so we still need to make it slightly fun to look at and engage them.
Then let's tackle the biggest issue, and make the y axis go all the way to zero and adjust the country data points so they reach the bottom
As we only have 6 data points we can scrap the axes and directly label them to reduce the clutter further. We can add the metric values too to appeal to a wider audience.
Lastly let's change the title to better inform the audience better as to what they are looking at. Ironically, titles are often overlooked by data viz creators, even though it's the first thing their audience will look at due to visual hierarchy.
And there you have it, we've gone from this particularly bad viz, given it a renovation and learnt a few data viz tips along the way.
If you have any bad vizzes you think could do with a renovation, let me know by clicking on the link in the description below.
Follow Datasaurus-Rex here:
Facebook ► http://bit.ly/2VKVbvL
Twitter ► http://bit.ly/2IXf6Gd
Instagram ► http://bit.ly/2XKXp03
Email ► http://bit.ly/2EUBeNr
LinkedIn ► http://bit.ly/2IWyrHC
YouTube ► http://bit.ly/2IWyvXS