Never have you spoken a truer word. The question I'm asking, is why doesn't it exist? It used to!!!
As regards your attack against the use of averages. It barely merits a comment so dim it is.
Think about it. You're saying that you can't use averages becauses data sets are made up of multiple different values - wow wee - you don't say!!! And yet with no sense of irony, you're happy to invoke your own next door neighbour as a proxy indicator instead (a sample of one) admittedly with no variation in the dataset, I'll give you that much anyway!
You have to start somewhere to try and get a picture Clive, and you're more than happy to invoke averages when it suits an argument you're trying to make, or are you introducing a new concept in quantitative analysis; 'the selective average'. Your argument about Hastings is of course superfluous and just an attempt to distract. Just to assist you though, I will help you by explaining how a fairer statistical investigation would approach that.
You would take the view that Hastings appears in the bottom decile (or research it to find out where it is using the SOA decile housing domain score), and therefore confine your search to the bottom decile for the corresponding salaries so as to ensure that you're comparing like with like. Your implied suggestion that we'd compare Hastings with Central London is simply bereft of understanding as how quantitative techniques are applied. In urban geography we have something called Super Output Areas which allows you to drill down to neighbourhood level, but you need an OCSI subscription to access it. So long as you're comparing like with like, it doesn't matter. Where on the scale you intercept it however does, although ideally you'd do so at each decile
Ideally you'd like to draw all the data from the same source, but this isn't going to be possible with the ONS as they're systematically refusing to present the housing data. As I've said, they've taken the averages and converted them into Z scores. This is a more lengthy process. They won't have done this for ease of collection and presentation, and they could easily have included the hard data in an attached spreadsheet, but they've systematically failed to do so. From the Z score they're introduced a meaningless index. It's a clear attempt to avoid presenting the bottom line figure, as the Z score is a product of the original data!!! They've clearly had it at some point to build the index
If you're so confident that "there is nothing to hide" perhaps you'll accept the challenge of finding it? I doubt you will somehow (I am sure it exists incidentally - or should be capable of being extracted) - but right now I can't lay hands on it
The opening sentance I posted in that post was
"I've been trying to model something along this by way of investigation" - basically if you were a fair minded and balanced indvidual (which you doubtless are) you would realise I don't know the answer and am therefore investigating it. I don't know what i'll find, but it would be a lot easier if the government weren't hiding behind Z scores and index's to try and conceal the figure
