Hard Graphed

We're all familiar with the staple Lib Dem electoral tactic of using distorted
graphs, but are some of our public bodies now starting to join in?
(Pic : via Wikipedia)

Yesterday, BBC Wales highlighted the findings of the latest Child Measurement Programme conducted by Public Health Wales (pdf). The programme weighs 4 and 5 year olds – around 31,000 – to determine levels of excess weight and obesity (see also : Fat of the Land : More work needed on childhood obesity).

The headline figures show that Anglesey now has the highest proportion of overweight or obese children (32.4%), closely followed by Merthyr Tydfil (32%) and Pembrokeshire (31%). Vale of Glamorgan (21%), Cardiff (22.6%) and Monmouthshire (23.3%) have the lowest percentages. On average, around 11.8% Welsh 4 and 5 year olds are obese, and there's a correlation between higher levels of obesity and deprivation. It also found there was little difference between rural and urban areas.

Putting these findings aside, BBC Wales have started to use infographics for Welsh political stories on their website to get key points across to readers - whether that's Assembly inquiry reports and reports like this. A picture is, after all, worth a thousand words.

One particular example is the graph on page 42 of the report (pdf), re-produced by the BBC, which compares obesity levels in the regions of England to both the Welsh and English average.

(Pic : Public Health Wales via BBC Wales)

Hmm.

Statistics can be presented in a deliberately misleading way to prove a point or to get the public's attention. There's no contention that obesity rates are higher in Wales, but this is a dodgy graph the Lib Dems would be proud of.


As you can tell - even taking into account the fact the bars could be based on a decimal figure, not a whole number (which would explain it, but still means it's misleading) - the bars for Wales, London and North East England have been exaggerated to such an extent it appears obesity and excess weight levels are more deviant from the norm than they are. It's the only graph in the whole report that's been distorted like this.

Similarly, it implies the gap between Wales and England is bigger than it really is. The bar for Wales is actually, based on a scale fixed on the English average and South East England figures, positioned at 29-30%, not 26% (the real figure).

This is what a properly scaled graph should look like based on the whole number figures provided :

Suddenly, Wales isn't doing quite as poorly compared to England and its regions.

Childhood obesity is a major public health concern, and arguably as big a problem as smoking and antibiotic resistance. I don't blame health authorities, or the media, for putting a report under the nose of ministers, other politicians or the general public that spells that out in as stark terms as possible and, perhaps, stir people into action. Appealing to, and reinforcing, the Welsh inferiority complex is one way to do it.

I still would've expected better from Wales' leading centre for epidemiology though.