March 28, 2013
Briefly
- Via @bradleyvoytek, a reasonably good article on common media/advertising statistical errors, from Cracked.com of all places (!)
- Where all the plastic goes (in the sea) (via several people on Twitter)
- An interactive infographic on accident blackspots at Stuff. I’m not really convinced by the calendar heatmap — I don’t think it’s that good for seeing patterns.
- How do you store and preserve Open Data: a story at Nature News (via Aimee Whitcroft)
- And since it’s a long weekend coming up: something that’s not remotely statistics, but is Quite Interesting. Siouxsie Wiles has another bioluminescence animation up on Youtube, on the Hawaiian bobtail squid, invisibility cloaks, and quorum sensing.
Thomas Lumley (@tslumley) is Professor of Biostatistics at the University of Auckland. His research interests include semiparametric models, survey sampling, statistical computing, foundations of statistics, and whatever methodological problems his medical collaborators come up with. He also blogs at Biased and Inefficient See all posts by Thomas Lumley »
Hi Thomas,
I did the calendar heatmap for the Blackspots interactive. I’m curious as to what could be a better way of visualising this data set to reveal patterns. I’m looking to update the visualisation soon and any suggestions would be helpful.
Cheers.
12 years ago
I don’t know what would be better, but I would have thought you’d want something that focuses on day-of-the-week, seasonal, and perhaps time-of-day patterns. The heatmap has some ability to pick up day-of-week patterns (Saturday looks bad), but not so much for seasonal patterns. Perhaps line graphs?
The heatmap is good for emphasizing the general background variation, which is something that headlines often mask.
12 years ago
Thanks. I’ll see if I can make it better either by using line graphs, or by providing more detail when the readers hover over the heatmap.
12 years ago
I thought it was interesting (the car crash chart). I hadn’t known about some of the deaths in my area. It would have been interesting to know if they were pedestrian-automobile crashes or am-am crashes.
When you zoom out and you start getting big numbers, it might be more useful to go to rates rather than absolute numbers (and change from (x,y) position of crash to number of crashes in a meshblock, area unit etc). But then you have a problem with “what’s the denominator” – you might be able to get some traffic volume for meshblocks from the MoT or, not quite so good, people counts from StatsNZ.
12 years ago