Create your own fine print
Various ways to show variability

Stutter steps, and functional legends

Dona Wong asked me to comment on a project by the New York Fed visualizing funding and expenditure at NY and NJ schools. The link to the charts is here. You have to click through to see the animation.

Nyfed_funding

Here are my comments:

  • I like the "Takeaways" section up front, which uses words to tell readers what to look for in the charts to follow.
  • I like the stutter steps that are inserted into the animation. This gives me time to process the data. The point of these dynamic maps is to showcase the changes in the data over time.
  • I really, really want to click on the green boxes (the legend) and have the corresponding school districts highlighted. In other words, turning the legend into something functional. Tool developers, please take notes!
  • The other options on the map are federal, state and local shares of funding, given in proportions. These are controlled by the three buttons above. This is a design decision that privileges showing how federal funds are distributed across districts and across time. The tradeoff is that it's harder to comprehend the mix of sources of funds within each district over time.
  • I usually like to flip back and forth between actual values and relative values. I find that both perspectives provide information. Here, I'd like to see dollars and proportions.

I also find the line charts to be much clearer but the maps are more engaging. Here is an example of the line chart: (the blue dashed line is the New York state average)

Nyfed_linechart

After looking at these charts, I also want to see a bivariate analysis. How is funding per student and expenditure per student related?

Do you have any feedback for Dona?

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

derek

The pauses in the animation seem like a good idea, but aren't they only an improvement in so far as they make the animation more resemble a small multiple? If so, wouldn't it be better just to go for the small multiple?

I quite liked the stuttering at the individual polygon level, though. It made the more fast-changing polygons more eye-catching.

The comments to this entry are closed.