For week 12 we asked the community to work with data about the UK’s pet population.
The dataset was pretty simple and there is a good reason for it. For all the variety of topics and challenges we want to provide, sometimes we just need to keep things very basic for our own benefit. With Andy completing yet another marathon last Sunday and me enjoying my first proper holiday in quite a while, I wanted to make sure I could create a makeover very quickly and be done with it. Being away and returning to Germany on Wednesday night only to leave my place 5 hours later to go to an event for work also meant I didn’t have a lot of time to engage with the community this week and provide feedback beyond our Viz Review discussion.
But there is always the recap, so let’s have a look what happened this week, now that you know why things have been a bit quiet and low key :-).
As expected, we saw quite a few bar charts comparing pet populations across different types of animals. Bar charts are a great way to compare categorical data, are easy to create and easy to understand.
We also expected to see icons used because the topic of pets is a playful one that lends itself to a more ‘fun’ style of visualizations and it’s good to experiment with different designs to see what works.
I was pleasantly surprised that a number of you turned this simple dataset, which – let’s be honest – didn’t have great depth for amazing insights, into thought provoking data stories by doing additional research and using data from the annual report of the PFMA to enhance your visualizations.
LESSON 1: CREDIT YOUR SOURCES
What we see very often when icons are used is that the authors fail to credit the source. Unless you create the image yourself, use your own photo, design something from scratch, or paid for it you must always credit the source of your image. And you should only use it when you have the right to do so.
That means if you’re searching for cute pet pictures on google images, you should filter using the license settings to only return images that are free for reuse.
Imagine if someone used a picture of your pet, child, spouse, house, etc. in their work without asking you for permission first. Feels wrong, doesn’t it?
Please do the right thing and don’t use icons, images, photos and graphics that you don’t have the right to reuse. We’d hate for someone to actually get in trouble for it and as your work is published online for everyone to see, you are exposing yourself to the risk of someone actually getting upset (or worse) about it.
While I call this a ‘lesson’, most of us have made the mistake of using an image without giving due credit. I just want to make sure everyone is aware that this can actually cause problems and to keep it in mind as a best practice to adhere to.
LESSON 2: USE LABELS TO INFORM, NOT TO CONFUSE
We had a few examples this week where both metrics, the number of pets and the percentage of households with those pets, were used in the same visualization. What was confusing was the use of labels in these visualizations. Let’s look at an example:
In this colorful viz, a lollipop chart represents the percentage of household with the pet as the length of the bar, while the number of pets in millions is the label at the end of the bar.
— Ravinder Singh (@RavSinghK) March 19, 2018
What this results in is a bar chart where the bar with the label 54 million (for any pet, here in red) is a little less than twice as long as the next bar, which is for dogs, with the label 8.5 million. That is seriously confusing: Why would a bar of 54 million be twice the length of 8.5 million? It should be 6.35x as long…
The reason is that the length of the bar is the percentage of households with the pet. That is indicated by the x-axis but is not explained in plain English anywhere.
So what the audience sees is a bar chart with confusing labels and below it there is a row of percentage number with seemingly no connection to the chart.
Sure, we have worked with the dataset and know what it is about and maybe to the author this was a very logical way to present the data, but to an uninitiated audience it will be very confusing.
So please, when you add labels to your visualization, take a step back and ask yourself whether the labels are in the right place, whether they add value rather than confuse your audience and whether they make sense given their context, i.e. the labels around them.