## We sleep HOW MUCH?

by

I’m an alum of UW-Madison, and have many fond memories of the place (indeed, TwoPi and I met in grad school there).   As a bonus, they have a good alumni magazine.  But even the best magazines make mistakes from time to time, and  the Spring 2009 issue of On Wisconsin includes a graph that is simply wrong.

The graph, illustrating how much time we sleep compared to other activities, appeared in the article “Bedtime Story” (description “Why do we snooze?  UW researchers are pursuing an unconventional theory — that our brains downsize while we sleep, getting us ready to face another day.”)  Here’s what the graph looks like:

Take a look at the two biggest circles:  we spend 36% of our life sleeping, and 19% on other [unlisted] activities.  But that gray circle is pretty small — you could fit 3 or 4 of them into the Sleeping circle.  I grabbed a ruler and, sure enough, its the diameters, not the areas, that have the 36:19 ratio.  The distinction is even more striking when you look at the tiny circles:  we spend 36 times as much of our time sleeping as we do on telephone calls, mail, and email and that’s a LOT, but it’s not a ratio of 1296 (as the areas would suggest).

So I created a graph in which the areas are in correct proportion:

There’s still a clear difference in how much time we spend doing different things, but it no longer looks like 61% of our lives are spent in bed.

Tags: , , ,

### 9 Responses to “We sleep HOW MUCH?”

1. samjshah Says:

Hahaha, my alumni magazine did the SAME THING a while back.
http://samjshah.com/2008/07/11/mit-im-disappointed-in-you/
Ah, statistics.

Sam

2. sumidiot Says:

Rather like this page from “Good Math, Bad Math” just yesterday:

http://scienceblogs.com/goodmath/2009/03/more_stupid_graphs.php

3. Sean Says:

I just saw another post with similar complaint against graphs earlier today, but I’ve already forgotten where it was. It had to do with state unemployment rates, though, if that triggers a bell.

4. Kate Says:

I think Sean’s referring to this post at Good Math, Bad Math.

It seems like circles is the new fad in infographics these days, and people keep trying to represent data with diameters, but the visual effect is that the viewer is comparing areas, with skewed results. Unfortunately when most people hear the word “diameter”, their eyes glaze over.

5. Ξ Says:

Thanks for the links to your site, Sam, and to to Good Math, Bad Math’s post yesterday: I had no idea this was a recent trend/fad.

[Incidentally, sumidiot posted the link first, but the first time people comment we have to approve it so it didn't appear until after Sean and Kate posted.]

6. sumidiot Says:

First!

Of course, I saw it from Kate’s delicious links (which looks odd to write…)

7. Dave Richeson Says:

A classic blunder!

Here’s a similar goof from the Financial Times. They fixed it after many readers pointed out the diameter/area problem. Here’s the fixed version.

Here’s the original article. Apparently the graphic came from a JP Morgan report; and JP Morgan just happens to look good with the misleading infographic. Coincidence?

8. Sleep Notebook Says:

Glad to see the sleep circle is actually a big chunk of the pie in the end. Seems like sleep is becoming less and less of a priority as our world gets more hectic and there are more technologies to occupy our time when we are awake. I am fighting to get more sleep than I ever have in my life and to stick to a routine. I am tracking my sleep in a blog diary I like to call sleepnotebook.com which helps me see progress and downsides to sleep, and share my knowledge with others as I learn, come check it out.

9. Jim Wolff Says:

Nice post.

I’m trying to discover the name of the type of graph you used to illustrate this data (so well). Does it have a name?

Thanks, Jim