Ignore visual metaphor at your peril

I went to a research talk recently. It was a brilliant example of how to talk lucidly about science despite having poor graphical representations of the data.

The the speaker was youngish, eloquent, and unashamed of it (more people should really use the word perforce). He even referred explicitly to the narrative: “The story I am going to tell you today …” So he had me on-side from the start, because I am a sucker for narrative, and the storytelling aspect of scientific communication.

Unfortunately, his graphs and tables and figures were pretty bad. Most of the time it was all I could do to figure out what the data actually were actually saying. About halfway through his presentation, I more or less stopped looking at the slides because I couldn’t understand them, and trying to figure them out was making me miss what he was saying.

So here’s an example of how not to illustrate the relationship between two variables, A and B, under conditions 1, 2, 3 and 4:

mean-difference_bad.jpg

Straight away, we are having to work harder than we should, because the scale is inverted, and doesn’t follow the conventional metaphor of “up is more, down is less”.  (Yes, I understand that negative scales should go downward, but when you are emphasising increases in the difference between two variables, it’s just clumsy presentation.)

What he could have done instead was this, in which up and down retain their metaphorical values of ‘more’ and ‘less’, respectively:

mean-difference_better.jpg

Here, the x-axis crosses the y-axis at zero, and that the magnitude of the difference between A and B is the same, but that I’ve ditched the negative scale. Why? Because it’s just more intuitive to think in terms of positive values: we do this every day when we handle pieces of fruit, or money.

Also, I just don’t find ‘mean difference’ graphs all that intuitive. Why show people the results of a subtraction when they can work it out by eyeballing the difference between variables, like this?

mean-difference_good.jpg

(This last figure is also more information-rich, since it retains the absolute values of A and B.)

I didn’t have enough ego to approach this guy and say “listen, this stuff is great, but your diagrams kinda suck.” Maybe I should have — sure, I’d have been more diplomatic than that — but I didn’t feel like I had the moral authority. Plus, he was a very good speaker, and it would’ve been rude.

Einstein allegedly wrote that “The supreme goal of all theory is to make the irreducible basic elements as simple and as few as possible without having to surrender the adequate representation of a single datum of experience.” Maybe we shouldn’t be in such a hurry to surrender those experiential data.

Advertisements

2 Comments

Filed under other people's stuff

2 responses to “Ignore visual metaphor at your peril

  1. Pingback: You need to love your audience (at least a little) if you want them to love you back « Finite Attention Span

  2. Pingback: Beauty, abstraction and storytelling « Finite Attention Span

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s