The Thanet Times (Tuesday, October 31, 2006) wanted to shock us with the "fact" that we have 119 registered sex offenders living in Thanet with a population of 120,000.

Then without converting to a compariable measure they state the Kent average is 59 per 100,000.

Matt looks at the numbers and shows us what they mean

These numbers need to be scaled. The number per 100,000 needs to be given as a number per 120,000 or Thanet's number per 100,000 needs to be given.

Let us look at those numbers as percentage of the population.

It's just 0.00059% of the population are sex offenders and 0.000992% of the population of thanet are.

If you got 1000 random people into a room odds are you would not have a registered sex offender in that room.

In fact if you shake hands with 10,000 randomly selected people there would be just a 9% chance (less than 1 in 10) that you would have shaken hands with a single offender.

To give you some scale you would need to shake one hand every three seconds for just under 8 and a half hours without a rest to have reached a probability of less than 1 in 10 (9%).

The entire readership of the Thanet Times is 35,000 which means there is less than a 32% chance that any sex offenders even read the Thanet Times!

Population of Kent in 2001 was 1,329,653 and the rate of increase has been VERY small due to the fact most people leave thanet as soon as they get a chance.

...so based on nothing more than the average we can only approximatly guestimate that there might be 785 Registered sex offenders in Kent. Given that the average is a meaningfull figure (which, you will see, it is not).

That means according to those figures we would have 15% of the sex offenders of Kent in Thanet but the Thanet Times suggests we have less than 12.5% given that 12.5% is supposed to be shocking.

So what can I tell from this?

That if 12.5% is more than 119 (which the Thanet Times suggests it is) then there is at least one more area in Kent with vastly more sex offenders than Thanet.

(or the figures from the paper were made up on the spot).

It is likely that there are a few "lucky" areas with significantly less offenders than the rest of Kent and that these places are causing a squew in the figures.

The upshot of this is that almost all of Kent can cry "good hevens we have a greater than average number in our area!"

The average is 59 - What could cause this number to be so low compaired to us?

Possibly there was an area that rated just 8 per 100,000 or some other very low number (probably the villages).

If that is true there must also be other areas with a higher than average number of offenders. It also indicates that there are areas with a higher count per 100,000 than Thanet.

It shows that the deviation from the mean is relativly large (no one has anywhere near the average). Therefore the figures given for comparison are meaningless scare-mungery on a slow news day. (As if the headline to story ratio didn't give that away.)