I’ll admit, life has been busy lately! Some of you know that I have two young boys (ages 5 years and 11 months) that keep me crazy. I mean, I’m probably a little crazy anyhow (like that time I recorded myself in the Las Vegas hotel at TC15 lip syncing to “watch me viz, show me data”), but keeping up with their energy makes me a little nutty!
All excuses aside, I owe everyone a summary of June’s Project Health Viz!!! A BIG heartfelt thank you to Anna Dzikowska, Nicole Lohr, Young Song, and Sean Ramsdale for sharing your take on some interesting healthcare data!
Just to reiterate the purpose of this project – the goal is to increase the visualizations of healthcare data and to gather healthcare data sets for anyone to use. Healthcare takes many forms such as personal well-being, information about hospitals and healthcare systems, global disease prevalence rates, statistics on states with the the lowest obesity rates, average number of cups of coffee people drink, the cost of the burden of disease, health risk factors, and everything in between! It’s anything related to our health and wellness. Remember the recent Iron Viz competition? That was timely – and it’s just an example of the variations on health and healthcare data that are available. You can view what people submitted here.
Data as Learning
In June (month two), I shared data from the World Health Organization on childhood mortality. Having two children, I was curious what were the leading causes of illness, and unfortunately death, in children around the globe.
What I love about projects such as this, Makeover Monday, and others is the opportunity to learn something. For those that work in healthcare, you may think, “I don’t want to viz on healthcare when I do that all day long.” That may be true (though I’m also hoping people not in healthcare use this as a growth opportunity), but, at least for me, every data set is a new adventure and a new subject to learn about. If you DO work in healthcare, then this project may be a way to increase your own Tableau Public healthcare vizzes since it’s highly likely you can’t share all your hard work you do every day with the rest of us! This means your Tableau Public profile can be used as a resume of sorts. Eva Murray says this in her blog: “If you build stunning Tableau visualizations and no one sees them, do they have an impact on the dataviz community?” Exactly.
Choosing a Dashboard Type
So what was my approach to this data set? At work, I am often asked to create one of two types of dashboards: an interactive, exploratory one, or a static, printable one. I’m a big fan of Adam McCann’s types of visualizations blog post where he explains, according to him, the three types of dashboards: data art, infographic, and dashboard. On Tableau Public I often try my hand at each of these depending on the story I want to tell or the skill I want to practice.
(Image from Adam’s blog and posted with permission)
For this month’s data set, I wanted to do something in between an infographic and a dashboard – something interactive, but something that could also stand alone as an infographic. That means that the story can be interpreted in just an image, but there may be some interactivity for someone using the actual Tableau version.
It’s All in the Details
My approach to designing a dashboard is often fluid. I don’t always start in the same place. For this one, while I did play around with charts first, I actually went to the internet and looked around for a color scheme. I typically Google infographic images or look through Tableau Public for visual inspiration if it doesn’t come easily to me.
“My approach to designing a dashboard is often fluid…I typically Google infographic images or look through Tableau Public for visual inspiration.”
All I knew was I didn’t want to go the red = death color route, which is ironic, since that is sort of where I ended up. But hear me out.
I my visual research, I recalled a viz I saw by Chantilly Jaggernauth on Ecological Footprints for Makeover Monday. I was drawn to the hue of red she chose and the simple shades of gray. I decided maybe I would use red, but instead of making everything red, I would use it as my highlight color and have everything else fade to gray. This helped me use color particularly to call-out certain aspects of the data both in the charts and in text.
Once I had my theme, I explored the data for my story and chart types. In the end, I used three chart types in this viz: bars, donut, and comet. I often use projects such as Makeover Monday or the Storytelling With Data Challenge (SWD) to try new chart types or new techniques that perhaps I haven’t used yet at work. What is helpful about this project it’s that I can try to apply these to healthcare data that may have some relevance to what I do at work. I’ve certainly used bars and donuts before at work, but a comet chart, no. So this was a chance to see what scenarios would work.
Bar charts are definitely one of the best charts out there. They are quite versatile in my opinion. I tried a line chart first for the number of live births over time, but found that when graphed it was hard to see the change since the total from 2000 to 2016 increased by only 10 million on a very large scale to begin with. Instead I wanted to show the change from each year. Using bars this way really highlights how despite the fact that the number of births still is increasing, the rate at which it has been increasing actually has been declining since 2006. It’s always best to try a few charts that help to tell your story best.
Donut charts are often frowned upon by some in the data community, but I have found use cases for them. Since I was already going for sort of an infographic feel to this viz, donut charts work. My point was just to visually show that these four causes of death took up half of the donut. By breaking it out into four donuts it makes the visual, from left to right, look like movement. Imagine a flip book and as you flipped, the red piece moved clockwise. Additionally, I wrote out the percent (the section of the donut) under each one. This was intentional because the point of the chart type was more for visual aesthetic than actual practicality.
Lastly, was the comet chart. As I alluded to above, I wanted to try out this chart with a healthcare data set even though I had done it before in other vizzes. Comet charts use two data points and the difference between them. In this case, I chose to highlight the difference in deaths from 2000 to 2016. It’s like a slope chart except I am able to plot each of the causes as rows.
You can see the interactive version here. Thanks!
Thanks to everyone who shared their #projecthealthviz visualization!! You can see all the dashboards below.
M2/Y1 Community Visualizations
Nicole Lohr’s Interactive Viz
Sean Ramsdale’s Interactive Viz
Young Song’s Interactive Viz
Anna Dzikowska’s Interactive Viz