What am I looking at?

We can work out the notional 'distance' between PCTs based on disease prevalence. If you imagine that we were looking at two areas and plotted a chart with the position of each PCT with each area on one axis. You could then get a ruler and see how far apart they are.

I have done the same here, only using 11 areas, which would give an eleven dimensional chart. The distances are called Euclidean distances. These form the basis of the "is like" section in each PCTs listing.

The next, and most clever bit is the squashing of that 11 dimensional space into something that we can see on a monitor. This is not exactbut is the best that can be applied. The maths program to do this and the java applet above that displays the data were written by the clever guys at The Public Whip and there is a rather better explanation on the MP map page of that site.

Colours have been added based on the socio-economic classification of each area as derived by National Statistics.

What does it mean?

This chart only shows distances, but you can see outliers from the main body. The colours also see what would be expected for each area.

Many times I have been told how our drugs budget compares to areas regarded as similar in socio-economic terms. This analysis give a new and potentially complementary technique to allow comparisons.

The Software

The viewer above and the maths program is released under the GPL. You can download it from here. However unless your needs are extremely similar to my own I would reccomend using the original as their project management is a lot better than mine!