Back to the Iowa Caucus results for a moment. A lot of the day-of forecasting for elections is done by entrance and exit polls. So in this piece from the Washington Post, we take a look at entrance poll results. This is basically a two-parter. The first is showing each candidate and the group they won and a number indicates by how much they won the demographic group.
Select the 30–44 age group
If you click on any of the demographic groups in particular, you are brought to the part of the page with the actual full results for the demographic. The format is simple a basic heat map with table. Nothing fancy, but nothing fancy is required for that type of data. Interestingly, the colour denotes not the share, but the result. I am not sure I would have done that, but it is a minor quibble.
The 30–44 age group results
Credit for the piece goes to Lazaro Gamio and Scott Clement.
It’s Monday, folks. And for most of us that means going back to work. Which means dressing appropriately. And that’s about as far as I’ve got introducing this subject matter, because I wear a dress shirt and tie everyday. Not a t-shirt. But we’re talking t-shirts. Specifically their sizing.
Threadbase is a New York startup looking to do some cool things with data about t-shirts. But that requires having data with which to play. And they are starting to do just that. Their opening blog post has quite a few data visualisations.
Comparing actual sizes via a dot plot
The dot plot above charts the sizes by dimension for various brands and makes. I might quibble with the particular colours as the red and purple are a bit on the difficult side to distinguish. Symbols could be away around the issue. But the only real issue is that on my monitors the full image runs long and I lose the reference point of the actual dimensions in inches.
But the piece is worth the read for the cyclical changes in dimensions.
Mostly it’s just a pity that I’m not a jeans and t-shirt sort of guy.
Today we look at a really nice piece from the Washington Post on urban homicide. It combines big, full-width images that use interactivity to promote exploration of data. But as you can see in the screenshot below, the designers took care to highlight a few key stories. Just in case the reader does not want to take the time to explore the data set.
The growth rate is an interactive piece
But the piece uses scale to provide contrast throughout the article. Because in addition to the three or four big graphics, a similarly well-thought-out and well-designed approach was taken towards smaller, inline supplemental graphics. Here is an example about the homicide rate for New York.
New York’s homicide rate as an inline graphic
What I really enjoy about these small graphics is the attention paid to highlighting New York against the background averages provided for context. Note how the orange line for the city breaks the grey lines. It is a very nice detail.
Overall, this is a really strong piece marrying written content and data visualisation.
New year, new selfies. Thankfully we have the Selfiecity to look at a sample of selfies, the goal to determine patterns and trends in the art of the selfie. Of course you also want to be able to look at the aforementioned selfies. For that they built the Selfiexploratory, an application that allows you to filter the set to see what you want. What I like is the return of data to show what the results of the filter look like against the whole set.
Last night, in the States’ time at least, North Korea purportedly tested a hydrogen bomb. How does this differ from their previous tests? Well, those were all nuclear fission bombs, this is a nuclear fusion bomb. (Admittedly, I am simplifying a lot here.) Hydrogen bombs, the H-bomb, are more powerful and more efficient in that they emit less radiation. They are still pretty bad news, though. That bit has not changed.
Anyway, the Washington Post put together a nice piece about nuclear weapons testing. The big feature piece is a map of test sites over time. What I really like about it, however, is that they chose to split the world at a different point—the Pacific Ocean opposite the Prime Meridian. I have occasionally argued for using such maps more often given the increasing relevance of Asia and the relative decline of Western Europe. So it is nice to see it put to good use here.
So yesterday we reimagined a less-than-stellar BBC chart. Today, we look at a good chart from the BBC about climate change, timed to coincide with the start of the Paris climate talks. This comes from an article with six charts related to climate change, but it is the best in my mind.
The trend…not so good.
Nothing but nice design here with the use of colour to highlight the top ten hottest and coldest years over the last 225+ years. But it really comes alive when animated and tells the story how those coldest years occurred at the beginning of the set and the hottest are among the most recent years.
Credit for the piece goes to Emily Maguire, Tom Nurse, Steven Connor, and Punit Shah.
You may recall a year and a half ago a post I wrote up about a New York Times piece looking at the fandoms of baseball in the United States. Well fresh off their hometown Royals’ World Series victory, the folks at the Kansas City Star revisited the graphic—driven by Facebook likes—to see if there had been any change. Sure enough, Royals Nation—or whatever they call it—has made inroads into what was before St. Louis Cardinals territory.
Missouri is a little more blue these days
The only sad part about the article is that they talk of changes in adjacent states, e.g. Kansas, but have no maps for those.
At the end of the month the world will gather in Paris, France for the next round of climate change talks. In advance of the talks, the Financial Times put together this model of how emissions reductions will help—or not—get climate change under control. The piece is two-fold. The first is a ten-step narrative that showcases the tool’s split of the time series into short-, medium-, and long-term impacts and how those work in the best and worst case scenarios. But it then allows the user to jump right on in and create their own scenarios.
Is it getting hot out there?
Credit for the piece goes to John Burn-Murdoch and Pilita Clark.
Canada held an election yesterday. For your briefing on it, John Oliver did a great job on Last Week Tonight. But for the serious coverage, we have results.
Here we have the results coverage by the National Post. It’s your standard choropleth coloured by the victor in each riding, or constituency. From a design side, I find the pattern fill interesting and not something I have seen done before for a political map.
I just chose a place I had visited in Canada
But I really like what the CBC did. They built an interactive application to cover the evening’s results as they arrived. This screenshot is for the riding in Fredericton, where my ancestors lived in the 19th century. (I had to have a connection to the ridings somehow.) In particular, I liked the ability to star ridings of interest and have them immediately retrievable. The CBC complemented that with a list of ridings to watch. It was a great resource for the evening.
But then they also covered the results with an article with interactive graphics. This is more your standard fare with choropleths, bar charts, and line charts. But they flow through the article quite sensibly. Overall, a solid results piece.
Party results per region
Credit for the National Post piece goes to the National Post graphics department.
Credit for the CBC piece goes to the the CBC graphics department.
Like the title said, it’s about time I took you all to school…by which I mean university scorecards from the US Department of Education. I used my alma mater, the University of the Arts, to show the design here. Basically you have several sections key to understanding a university from the student body to the financials to the graduates’ prospects.
The scorecard for UArts
Credit for the piece goes to the US Department of Education.