« Progress and retrogress | Main | Statistical science fiction »

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341e992c53ef00e551e915f18833

Listed below are links to weblogs that reference Cram it like Koby:

Comments

Jon Peltier

'PS. Anyone knows what is being measured on the horizontal axis? The original graph mysteriously stated "respondents' views".'

I spent a while looking at the original chart and the redrawn charts trying to figure out what was being measured. It wasn't until later I noticed your question, and realized that you don't know either. I went to the Economist article, and it wasn't too clear either:

'The data are derived by subtracting left-wing answers from right-wing ones, for each country and for each main political grouping within each country. A net minus rating suggests predominantly left-wing views and a positive rating suggests a preponderance of right-wing views.'

Clear as mud. What are left- and right-wing answers? Sometimes it's easy, like 'Do you believe there is a God?', which has either a Yes or No answer. Sometimes it's subject to more interpretation, or at least interpolation, like 'The profit motive is the best spur to job-creation', which has a spectrum of responses between strongly agree and strongly disagree.

Keith Watson

Setting aside Jon Peltier's important point about the concept of what the heck is being measured on the horizontal axis, the junk charts version is a great improvement. I like the version with lines because it allows my eye to follow which group is which as I scan vertically.

Stephen Hampshire

While the Junkcharts version is much clearer, I do feel you've lost the representation of the size of the gap between "left" and "right".

The Economist chart could be clarified, and would be very good with 2 parties. Representing all 3 is what makes it hard work I think.

Alan De Smet

What's with the lines in the junk art version? Lines between points suggests a continuum, which this obviously isn't; it's six distinct points. For example, it's not meaningful to examine the point between Ideology and Military action. If it's just there to help identify related points, surely there is a better way?

Alan De Smet

Sorry, obviously ignore the above. I didn't see the final notes in the post when I made my point. Perhaps it was added after I composed it, and then accidentally let it sit for a while. Or perhaps I just failed and overlooked it.

Andreas

One information that the junkchart versions do not convey as clear - I think - is that the US voteres are much more divided then the UK voter. So maybe the observation that
"the U.S. boxes grab all the attention by virtue of them being wider (country being more partisan)."
is the whole point?

Kaiser

Andreas: you pretty much summed up the issue in different words! If the point was to show the difference between US and UK spreads, then all the extra information about 3 parties and 2 averages only serves to confuse things. The horizontal bars highlight the differences at the level of issues; they don't provide a good overview of the average difference.

Bob

It always strikes me as odd how much white space the junk charts "improved" graphs have. Very little of the space available is used to represent the relevant information, leaving the graphs looking undernourished, straining the eyes looking for clear conclusions from emaciated lines. The other improvements seem great as a general rule -- but only after working harder than seems necessary to find them.

Kaiser

Bob: good point. I think the fear of white space is what leads to a lot of cluttered, confusing charts. This is no different from creating slide presentations: a common advice is to focus on one key idea on each slide, and not to cram the slide with text.

Bob

Cramming it with text certainly won't help. Making the lines thick enough to make the main points easily visible, using color conservatively to differentiate conditions... this can make it easier for the eye to catch the key idea without any added clutter.

J

No offense, but I tend to agree with Bob's points above. The Economist chart is not great, but the redesign doesn't seem like much of an improvement.

The Junk Charts version would take up about three times the space of the original chart, but it removes information (the country ranges are lost in the small multiples, so it's no longer visually obvious that the U.S. is more partisan), abbreviates the labels, crams them in so that they run into each other, and adds meaningless connecting lines. Beyond those serious technical problems, the charts look completely unfinished, as if they were built in Excel in about ten minutes.

The redesigned charts would look like a printing error if they were to actually run in the Economist, so it's not really clear what the point of the exercise is. Assuming you were actually given three times the space in the magazine, is this really your ideal redesign, what you'd like to see on the page? Like many of the charts on this site, the redesign seems to completely ignore the space limitations of the printed page, along with legibility, color and typography, all of which are essential to conveying information. The only thing the redesign does seem to show clearly is a strong preference for scatterplots.

Kaiser

J: you are entitled to your opinion. For me, I do not believe that the purpose of the original chart was to show that "the US is more partisan". That is what our eyes pick up; it only shows the chart's weakness.

In fact, the title of the piece was "Anglo-Saxon Attitudes" and the subtitle was "Our polls show the two may have less in common than they think."

Further, the solution to limited space is not to cram more data onto the chart. The solution is to simplify the message, clean up the chart so that the message stands out.

J

Yes, I agree that's the solution. My point is that for once it would be nice to see this blog actually try to make a chart that is complete, that has been cleaned up, that has been executed well, and that does have a legible message that stands out without paragraphs of text next to it explaining what the message is -- in other words, a redesigned chart that could actually stand in the place of the original in a real publication. That would be nice to see, but I have yet to see it on this blog.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Working...
Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.

Working...

Post a comment

Marketing analytics and data visualization expert. Author and Speaker. Currently at Vimeo and NYU. See my full bio.

Book Blog



Link to junkcharts

Graphics design by Amanda Lee

The Read



Good Books

Keep in Touch

follow me on Twitter

Residues