Post by Robert ChungPost by Mark & Steven Bornfeld DDSPost by Dan Connellyhttp://anonymous.coward.free.fr/temp/bushjobapproval.png
http://anonymous.coward.free.fr/scpo/exp-percap.png
http://anonymous.coward.free.fr/scpo/pct-gdp.png
http://anonymous.coward.free.fr/temp/fat-vote.png
These are simply amazing!! Is the anonymous coward one of you two, or
is he/she truly anonymous?
There's a little bit of anonymous coward in us all.
The plots themselves aren't that amazing -- they're pretty simple. What's
amazing is that so few plots are drawn in a way to make relationships
clear. Well, maybe not amazing; appalling is more like it. Most plots you
see in the newspaper tell you what happened but they hardly ever make you
wonder why it happened.
The idea for the Bush Job Approval plot was shamelessly stolen from
http://www.pollkatz.homestead.com/files/image003.gif
(since Pollkatz is no longer updating his plots) with the addition of
direct labeling; this is why Dan was able to pick out the consistent
pattern differences between Zogby and Fox. The Pollkatz plot is good
because it tells you what happened in a way that hits you between the
eyes--but see if you can spot the Zogby-Fox difference. The direct
"what's up with that?" Plots can't always answer "why?" but good ones will
make you curious enough to wonder.
You may have known that the US spends more than most countries on health
but the two plots of life expectancy vs. national expenditures make clear
how unusual the US situation is. Most plots that attempt to do this simply
show a histogram of per capita expenditures with the US way off at the
right, like this: http://anonymous.coward.free.fr/scpo/pct-gdp2.png. The
two plots listed above show context for those expenditures, and made Dan
think about how much bang we get for our healthcare buck. That's a good
question. BTW, note how direct labeling adds context that is lacking in
http://anonymous.coward.free.fr/scpo/pct-gdp3.png.
http://anonymous.coward.free.fr/temp/tdf97-bmi.png and
http://anonymous.coward.free.fr/temp/tdf.png)
use direct labeling to add data context or an extra dimension.
Thanks for this. Of course the healthcare spending/life expectancy
the right questions.