Light entertainment: bold paint flowing down
What's in a cronut? Let me find out

Sometimes you expose the holes

On Friday, I'm attending and speaking at the Leaders in Software and Art Conference, organized by Isabel Draves. LISA is an amazing gathering of artists interested in technology and software. For example, there is a panel on 3D printing and hardware hacking, and one on "creative coding, art and advertising". Check out videos from past years, and click here to register. My talk is at around 3:30 in a tightly packed day of activities.


Andrew Sullivan highlighted a chart showing the public attitude toward climate change globally:

Andrew summarized the above chart thus: "Sadly, America is home to far more climate skeptics than the global average."

This conclusion may be correct but the chart is less convincing than it appears.

Trifecta_checkupLet's pull out the Junk Charts Trifecta Checkup. Recall that there are three sides to the triangle. The question is well-posed, and the bar chart is an adequate choice for this data. We thank the designer for not printing the entire data set on the tight space, and to start the vertical axis at zero.

There are a few improvements one can still make to the bar chart. Start with turning it around so that the reader doesn't have to turn his/her head around. Also, extend the axis to 100% helps the interpretation a little bit.


If you have keen eyes, you notice that Greece showed up at the top of the revamped chart. The bar for Korea is also a tad too short in the original chart; it should be at 85%. 

To what extent is the set of countries "global"? Take a look:


It missed all of Scandanavia, most of Indochina, India, much of Africa, and all of Central America.

In the Trifecta checkup, we note that the data may not be complete for the posed question. Given this flaw, the map is perhaps a better choice to show us where the holes are.


Frank S

Well done. So much stuff published by those intent on alarming us about CO2 and climate is vulnerable to being exposed as misleading by anyone who digs into the data and the contexts. This is a very minor example, but more important ones include the MBH (1998) Hockey Stick (which took years to expose since the authors refused to share data and codes), the notion of a warming Antarctica, the disappearing glaciers of the Himalayas, the disappearing polar bears, the disappearing snows from Europe, the increase in hurricanes, the surging of sea level rise, the relentless increase in global mean temperature, the 'acidification' of the oceans, ... I am sure my list is incomplete.


Have countries been given equal weights in calculating the global average? I would expect the US to be typical if the mean were a weighted average based on population, given that China appears slightly more sceptical.


I think it would be better to have both charts. On the map my eye jumps to Brazil and Argentina, the third and sixth on the list. Greece and South Korea are more subtle on the map simply because of their smaller size (a known pitfall of chloropleth maps).

It's also not clear from the posting whether the data was truncated (the top rankings of a full data set) or a sample of countries, which does seem to at least be cross-cultural.


Frank: Your point is taken but my challenge to climate skeptics is to find their own evidence, data, and make their own charts. It's easy to criticize but you have to show the alternative, which is the spirit of this blog. A blanket statement claiming all observed phenonmena are natural is not evidence.

RobM: This one is tricky. Because the population size distribution is so skewed (China, Indonesia, India, etc.), any weighted average is meaningless. Computing a good set of weights for combining poll data from around the world sounds like a worthy project. Do any readers know any such weighting scheme (other than the predictable and predictably bad straight average, population-weighted average, GDP-weighted average, etc.)?

Henry: It's a sample of data plus I think either the designer or the charting software inadvertently left off the top. It is always true that multiple views of the data give a more rounded view but most publications can't afford to print both. However, maybe that's one of the rules that should be broken!

The comments to this entry are closed.