1 of 1 people found this helpful
It is a much better dataviz, with a lot of improvement compared with the original from The Economist Daily Chart.
Here is my my proposal,
I like my version because:
1.- It is a simple view
2.- A lot of information is in it. Two measures and two dimensions
3.- A simple linear regression model over imposed to the scatter plot provides additional analytic possibilities
The information it shows allows us to identify the amazing time in which the Sky City One skyscraper was built, but also to identify other outliers such as the Empire State building in NY, USA. In the other direction, it also allows to identify those skyscrapers that were built taking more time than expected.
Geoff - I think you're "Days/Metre" bar chart is the best option. The point of the article was to emphasise how much of an outlier the SkyCity is, and with your metric you've made that the biggest bar, by miles. For me - this answers the question perfectly.
I don't like the trend lines on the upper chart, though. Why? I don't think the two measures have any real correlation. Time and height are only 2 of many other measures that influence things (money, location, construction style, geography, etc, being some of the others).
Ramon - I tried a scatter plot too in my attempt, but discarded it. Why? Well, first of all I'll be clear - I agree with your 3 reasons for liking it. So what's my problem?
- SkyCity is buried along an axis. Even though it's an extreme outlier, it's proximity to the axis means it doesn't stand out.
- Scatter plots are hard for audiences to understand. This is a challenge for us experts. Should we use scatterplots often and hope people get educated enough to be comofortable with them? Or stick with bar charts with multiple encodings?
Great work, though - I love the way Tableau Public can be used to evolve and develop ideas.
Also - I rushed my example while travelling home yesterday; man, there's so much more we could do!