« The eyeball test | Main | New York Times: a tribute »

Comments

Ken

I assume that the type of each square relates to the group that is most over represented for that area, which is not a good summary. The first question should have been: Is there a single dimension ? Maybe there are 5 dimensions and it should have been 5 graphs. Or two dimensions: whether people are more or less likely to get married, and whether they are more or less likely to divorce once married. Age distribution would need to be taken into account, something that seems to be ignored in the original.

derek

Ken, no, the real clue comes when you realise there are black boundaries between some groups of rectangles. I *think* I can count 11 groups, and you're supposed to note the prevalence of the different reds within each group. It's our old friend the square pie chart!

PS Kaiser, please don't refer to the British as "Brits"; you wouldn't refer to the Chinese as "Chinks" or the Japanese as "Japs" in the context of a serious article.

derek

I see Danny Dorling is Professor of Human Geography at Sheffield University, and he ought to be ashamed of himself. Here's an equally insane map of the same data, one of a series of similar maps. And Ken's right and I was wrong: the maps have a key that says "Most unusually large group is shown when each place is compared with GB".

Those wishing to show Professor Dorling how it's done can download the Excel file of raw data from that page.

Peter H

Someone apparently made their graph in Excel, hence the squares...

Daniel Pope

derek, unlike 'Chink' and 'Jap', 'Brit' is not a pejorative.

ZBicyclist

This is truly a mess. Each block is 500,000, but what, then, do the blocks represent? Not Kent versus East Anglia, for example, because that would require knowing how these geographic areas mapped into the blocks.

And what about those remarried people. Do they really prefer the coasts (possibly because this allows escape by sea if this one doesn't work out)? Or is it some form of artifact.

A truly junk chart!

derek: Maybe it's a square pie chart, maybe not. I'm looking at the bottom center area, which is

M
S
S
MM
RM

If it's a square pie, shouldn't those M's all be grouped together?

ross

I think you're being a bit harsh, although I agree it not a brill chart.
Hi Here are my thoughts:

>>I assume that the type of each square relates to the group that is most over represented for that area, which is not a good summary

When based lined agianst average UK, we guess, I agree not very clever or usefull!

>>Age distribution would need to be taken into account, something that seems to be ignored in the original.

I dissagree with this as age is a dependent variable

>>I *think* I can count 11 groups, and you're supposed to note the prevalence of the different reds within each group. It's our old friend the square pie chart!

I dont think it's a square pie chart, but what is intresting about the "regions" is that they dont really match any UK regions - strange that they should be there, dont seem to offer much?

not a fab chart, but i can see what he's trying to do!!!


The comments to this entry are closed.

BOOTCAMP SUMMER '19



Link to Principal Analytics Prep

See our curriculum, instructors. Apply.
Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR.

See my Youtube and Flickr.

Book Blog



Link to junkcharts

Graphics design by Amanda Lee

The Read



Keep in Touch

follow me on Twitter

Residues