iitypeii

iitypeii is the work of two current affair nitpickers ([C] and [M]) and guests ([G]) giving their perspective and insight on publicly available statistics, the connections (and lack thereof) between them, and the erroneous conclusions made...

BBC - Resorting to Tabloid Headlines... \ [M]

The BBC resorted this week to the following eye-catching, tabloid-like, headline:

Of course, the key phrase here is "in parts of": taking enough subsets of the population will of course throw up some outliers which on the surface will look worrying. Comparing the worst subsets of the UK to entire countries is misleading and is like comparing apples and oranges.

Digging into the article a little more it turns out that to find the eye-catching figure they had to look at ward level data. In the UK a ward is a small geographical area with on average a population of 5,500. This of course means the headline is particularly misleading as in order for an average ward to have a incidence rate of 150 TB cases per 100,000 people, in actuality there would need to be between 8 and 9 cases. There is a certain amount of randomness in contracting TB and so comparing two wards which were otherwise identical you could realistically expect differing numbers of cases. Indeed, with such small case numbers this could be down to one particularly unlucky family. 

Further into the article there is a more informative graphic (albeit on a completely different size of sub-division of the population):

The article goes on to give public health reasons (such as chronic ill health) and demographic reasons (such as numbers of foreign-born inhabitants) as to why there are may be differences in prevalence between boroughs. Indeed it is entirely plausible these may explain the differences, but I'm less than convinced by the analysis. One obvious question to ask would be if we were to assume contraction rates across the UK for any individual (or family) were identical, what sort of natural variation of incidence would there be if we were to exhaustively search for sub-divisions of the population? What sort of variation in prevalence is there in other countries? All questions which go unanswered...

"If it cannot be expressed in figures, it is not science, it is opinion."
Robert Anson Heinlein