Saturday, 27 October 2018

Has the BBC fact-checked this?


Someone - namely Manchester Uni's Prof. Louis Appleby - has been fact-checking the BBC again and found it wanting.

  1. Tragedy of young life lost but BBC London News ran it under headline that teenage suicide rate in London is 4x higher than rest of country, repeated in story. In fact, rate in London is lower. Astounding piece of misinformation.
  2. How did this alarming claim, which must have been seen by countless people inc teenagers themselves (there is concern that stories like this can be self-fulfilling) come about? Answer shows how media can misuse stats & feed off each other.
  3. First it's worth stressing that suicide in <20s has been going up in last few years, a tragic situation. But that's all the more reason to have accurate figures & responsible reporting. Press sometimes seem to be competing for latest "mental health epidemic" headline.
  4. Story started in May when ONS published response to an FOI request about number of teenage suicides in borough of Brent, whole of London & nationally. Request was for 3 financial years (that's unusual), starting with 2013-14 . …
  5. As a result The Independent ran story of London suicides up from 14 to 29, calling it a "107% rise", >4x higher than 24% nationally. Figs for Brent were too small to mention but local charity that made FOI request was quoted, calling findings "shocking".
  6. But what's wrong with that? It is a 107% rise. Here are 3 tips on making most of stats: 1. Choose low baseline that makes any subsequent rise look far bigger 2. Go for short time periods or risk ruining a dramatic trend 3. Don't worry about small nos, they give you big %s.
  7. In their story BBC London News added a twist, misquoting The Independent's claim that the *rise* in London was 4x higher than nationally, a dubious distortion in itself, saying that the *rate* in London was 4x higher, which would have been astonishing if true.
  8. Then there were 2 other things that anyone in the field could have told them: 1 quote rates, not numbers - rates are adjusted for population size. 2 ONS suicide figs are based on date of registration, not date of death, & when nos are small, that can matter.
  9. So what is happening to teenage suicide in London? 1. Far from being 4x higher than nationally, rate is not higher at all, in fact slightly lower. 2. But in common with teenage suicide across Eng & despite fluctuations, it has gone up. That's the story.
  10. And the moral? 1. Dramatic claims based on population data like suicide stats are likely to be wrong. 2. Media stories feed off each other - bogus figs keep coming back - so worth getting it right 1st time. 3. And even if the country is tired of experts, they can be helpful.
But this just in from Prof. Appleby:
Credit to reporter on this who contacted me yesterday & has removed “4x higher” line from online story. So often these errors go uncorrected.
Unfortunately, I'm re-watching the BBC video report at the very moment and though the headline no longer mentions the offending statistic the report itself still does, saying "Teenage suicides are four times higher in London than the rest of the country". The error, it appears, has only been partially corrected.

1 comment:

  1. Seven out of ten BBC reports contain factual errors. Of those four also contain unwarranted bias. From a statistical overview of the BBC by the regulator OfSOD.

    ReplyDelete

Note: only a member of this blog may post a comment.