Monday, December 19, 2016

Making Comparisons with Data Visualization #Analytics #13

Statista produce some great data visualization on a daily basis - today's shows the top 15 countries in the world for the number of PhDs (data from 2014). The USA comes out on top with a whopping 67,449 PhDs graduating in 2014, more than double the next country Germany. As I will have a new module on Data Visualization to teach next semester I have a renewed interest in how data are visualized. I also have an interest where data visualizations by the likes of Statista like these can be a little bit misleading. I'm not questioning the data - just how it is show in a bar chart. At a glance it would indicate that the USA is the smartest country in the world - here's Statista's chart:

Infographic: The Countries With The Most Doctoral Graduates | Statista
You will find more statistics at Statista.

The chart does not take account of population and therefore lacks proportion (Statista does not claim this - it is simply plotting numbers). While the USA has just over double the number of PhDs compared to Germany, its population is four times greater. A more useful graphic would be the number of PhDs as a percentage of each country's population (pop data from www.worldometers.info). When you do this a completely different visual (drawn in Excel) emerges:

(Globe background sourced from Wikipedia)

In my version, Germany, UK, France, South Korea, Spain, and Australia are all ahead of the US - in fact instead of ranking first, the US ranks seventh in my list. And don't forget - this is just one year's worth of data - the actual number of people in each country will be much higher than the percentages above. 

I tell my students to challenge data when they see it. Ask how reliable they are, what do they tell you, is there a different way to visualize and analyse, how best should the data be used? It's fun to draw interesting charts, but we need to be careful how we use, and in this case, visualize the data.

No comments:

Post a comment