If we want to assess how well children are doing, and identify solutions to challenges in child well-being, we need a better understanding of the physical, social, economic and cultural environments in which children live and grow.
Efforts to track child well being in New Hampshire have been challenged by the lack of a single, accessible repository of data that would allow policymakers and citizens to understand how New Hampshire children are doing.
My client, Spark NH, contracted with me to collate and analyze indicators of child and family well-being in New Hampshire, and create an online platform for anyone to explore the data. With the fabulous graphic design and web building skills of Holly Valero at Hollyworks, I created this system: Visualizing Child Well-Being in New Hampshire
We see this tool as providing a one-stop shop to practitioners, policymakers and all those interested in a better understanding of the contexts that both promote the healthy development of New Hampshire’s next generation.
This is not an online data warehouse, but an interactive platform built with a cutting edge data visualization product called Tableau. Tableau not only allows us to tell the story of child well-being in New Hampshire in a visually interesting way, but also provides users with interactive tools that allows them to explore the data more fully. The charts and graphs can be sorted, downloaded as images or PDF files, and shared via e-mail or on the web with a single click. The raw data can also be downloaded directly either as a text file or as a Tableau workbook for others who also use Tableau.
Scrolling through my Twitter feed this morning, I came across this:
The way the data are displayed makes it unnecessarily difficult to draw the conclusion asserted in that post. There’s a lot going on here, with colors and other features of the chart competing for our attention. E.g.,
- What do the different colors mean? Is it significant that 2 columns are shades of grey and others are colors?
- My eye is immediately drawn to the smallest and tallest columns, which also happen to be near the center, and beside each other. Also, why is that value label (66%) IN the bar while the others are on top?
- Why does the Y-axis just go to 70%?
- Is the total column necessary info?
[If you like, you can read the Foundation’s blog post associated with this data]
I plopped the data into Excel and crafted this revision:
Full disclosure, I am a huge fan of the Annie E. Casey Foundation, and it is (easily) arguable that advocacy for child and family issues would not be what it is without their Kids Count data. But in this particular case, the design of the graph they created does not help make the point they wish to make.
*From a framing perspective, I also question why it was deemed valuable to share these data in this way. By calling attention to the low number of single-parent immigrant families (implicitly suggesting that’s a good thing), do these data inadvertently reinforce negative cultural stereotypes about single-parent families?
Do the data reinforce stereotypes about racial and ethnic groups also portrayed here?
Those are important questions to ask when presenting data. It’s also important to consider whether these kinds of data presentations, by triggering stereotypes, make the conversation we should be having more difficult to have. Namely, with over 1/3 of all US children living in single-parent families, how well are policies and programs responding to the reality of life for families today?
Charts, that is. Blueberry? Keep at it.
I have been doing away with pie charts in client materials for some time. This morning, I stumbled on the most excellent explanation of why, from Cole Nussbaumer Knaflic in her most terrific new book, “Storytelling with Data.” She’s written quite a bit on this, and even given a talk called Death to Pie Charts. A woman after my own heart, for sure!
The main reason I find the pie chart an ineffective communication tool is that it takes too much cognitive work for people to assign a quantity to an area. I do understand that pie charts CAN be useful in giving a quick sense of “percent of a whole,” so I’m not calling for their total eradication.
But many of my clients use them like this:
YIKES! Right? And doesn’t the red 28% look bigger than the blue 28%? As Cole explains in her book, things closer tend to look bigger. AND, using 3D can make things even worse for a pie chart’s interpretability. 3D charts throw in even more angles and dimensions for our eyes to sort through.
And don’t go running out for donuts to solve this problem. Cole shows why the donut chart is similarly troubled. In short, Why ask your readers to interpret angles and areas when a simple horizontal bar chart would suffice?
Do purchase her book, but before it arrives, check out this blog post she wrote on Alternatives to Pie Charts. It’s incredibly helpful, and I think you’ll be convinced!