### Time for Private Eye to go back to school to learn some maths

The current Private Eye has a little snippet in the Street of Shame section making fun of The Guardian's use of statistics. Unfortunately, the know-alls at Private Eye can't do basic maths and end up shooting themselves in the foot:

Ho ho ho. So funny Private Eye, you got out a pocket calculator and did 52 * 83% and got 43.16 and laughed and laughed. And then 52 * 11% and got 5.72. Ha ha ha.

Jokes on you though. The Guardian is only showing figures to two significant figures. Remember significant figures from school? No? Too busy making fun of the clever boys?

Given that The Guardian said 83% (and not 83.00%) that 43.16 should have been rounded to 43 and given that we're dealing with whole people it doesn't even make sense to think of 5.72 as 5.7 (two significant figures) but rather 6.

Now, if you wanted to point out an oddity then you would have done the calculation one way to get 43 Asian Pakistani, 6 Asian other and 3 white British (43 + 6 + 3 = 52) and then you would have worked out the real percentages. Let's say to two decimal places: 82.69% Asian Pakistani, 11.54% Asian other and 5.77% white British.

Then you could have looked at the figures The Guardian reported and actually spotted a real problem. 82.69% has been rounded to 83%, 5.77% has been rounded to 6% but 11.54% has been rounded down to 11%.

There's probably a good reason to do that. Suppose The Guardian had correctly rounded to 83%, 12% and 6%, you would have been laughing about the numbers adding up to 101% (because of the rounding). Given that 12% of 52 is the same as 11% of 52 at this level of whole person accuracy, The Guardian did the right thing.

Please report directly to Mr. Gove for remedial maths!

Ho ho ho. So funny Private Eye, you got out a pocket calculator and did 52 * 83% and got 43.16 and laughed and laughed. And then 52 * 11% and got 5.72. Ha ha ha.

Jokes on you though. The Guardian is only showing figures to two significant figures. Remember significant figures from school? No? Too busy making fun of the clever boys?

Given that The Guardian said 83% (and not 83.00%) that 43.16 should have been rounded to 43 and given that we're dealing with whole people it doesn't even make sense to think of 5.72 as 5.7 (two significant figures) but rather 6.

Now, if you wanted to point out an oddity then you would have done the calculation one way to get 43 Asian Pakistani, 6 Asian other and 3 white British (43 + 6 + 3 = 52) and then you would have worked out the real percentages. Let's say to two decimal places: 82.69% Asian Pakistani, 11.54% Asian other and 5.77% white British.

Then you could have looked at the figures The Guardian reported and actually spotted a real problem. 82.69% has been rounded to 83%, 5.77% has been rounded to 6% but 11.54% has been rounded down to 11%.

There's probably a good reason to do that. Suppose The Guardian had correctly rounded to 83%, 12% and 6%, you would have been laughing about the numbers adding up to 101% (because of the rounding). Given that 12% of 52 is the same as 11% of 52 at this level of whole person accuracy, The Guardian did the right thing.

Please report directly to Mr. Gove for remedial maths!

*If you enjoyed this blog post, you might enjoy my travel book for people interested in science and technology: The Geek Atlas. Signed copies of The Geek Atlas are available.*

<$BlogCommentBody$>

Create a Link

<< Home