How many months makes a trend?
I was very wary at the start to draw hasty conclusions, but with 5 months of totalled data, we do seem to have something of a trend.
There is still much to learn though. What I don’t know for example;
Is there a seasonal/temperature variation, i.e. more crime in better weather and longer daylight hours?
But that’s just the point – the more data we feed in to the database, the more we can determine relationships and meaning from it.
We have reached these figures in the chart above – see breakdown on our National Picture page – by totalling the data stream for streets by Police Forces from www.police.uk/data and trust them not to go back and change it.
Initially, we were using neighbourhood crime totals but we had to stop because every so often – in several hundred cases – the police neighbourhood boundaries were redrawn, the neighbourhoods renamed and all the historical crimes were lost !
This way – using street – i.e. boundary free data – we can be sure that the historical data remains cast in stone.
Now back to the main point. My gut feeling is that 5 months isn’t quite enough but 12 months of data would really give us a very clear picture of what’s happening.