Blog Post

Using graphs to distort facts

Is this graph deliberately designed to distort the facts?

It’s part of a direct mail letter I received recently, attempting to persuade me that I’m more likely to be a satisfied borrower if my money comes from a small bank.

A quick look at the graph suggests that borrowers linked to the Big 4 banks are very unhappy.

Or are they?

Perhaps the letter’s message is built around the assumption that most readers glance at the graph, without thinking about it too closely.

At a first glance, the graph suggests that customers of ING Direct and Bendigo Bank are highly satisfied with their home loans. In contrast, NAB and Westpac customers are highly dissatisfied. Taken on bar size alone, Westpac customers are more than 10 times less happy than ING Direct customers.

But this graph’s y axis starts at 70%, creating a visual impression that’s false. The numbers present a different story.

In reality, for every 100 ING Direct customers, about 95 are satisfied. For every 100 Westpac customers, about 73 are satisfied. A difference, yes, but not a huge difference.

If the graph’s y axis started at 0, then a 46mm bar for ING Direct would equate to a 35mm bar for Westpac (the graph sent to me compares a 46mm bar for ING Direct with a 4mm bar for Westpac).

Many readers struggle to understand numbers and statistics, and it frustrates me when poorly structured visuals distort a message. It may not be false advertising, but it’s not far from it.

I showed this graph to my 12-year-old son, and I’m pleased to report that he described it as ‘misleading’ and explained exactly what was wrong. Thanks, Queensland education system.