Tuesday, March 16, 2010

Statistical Scaremongering based on faulty interpretations

Possibly the biggest hole in N.Z. media reporting and editorial skills is in the ability to make sense of statistics – or more conspiratorially – their penchant to substitute melodramatic rubbish for immaterial fact in pursuit of the “big” story.

The ODT of 15 March carried banner headlines in about 28 pt font about the shocking level of child crime we are now experiencing. In fact 719 children under 9 had been “apprehended” and dealt with in one way or another (not a kick in the pants and a big telling off you may rest assured). Some of these children were as young as four – I suppose to be fair, in the interests of balance the paper ran a story in the same edition (different story) about the oldest offender (97).

So 719 offenders. As of the 2006 census there were 567,897 persons aged nine years old or younger in N.Z. thus we have the stunning “problem” of  0.13% of under nine year olds offending – insofar as that can be measured by apprehension. Even given poor apprehension rates this number has to be eight times higher to even hit 1% – and 1% is hardly worth a story let alone a front page headline.

Pickings are clearly slim on the story front at the ODT! Is this new? Well we are not given any historical data – there probably isn’t much so fair enough. The Police spokesman though expressed “grave concern”. A Mandy statement – he would say that wouldn’t he – I suppose he has to, but it is no surprise that kids, lots of the little blighters are frequently toe rags, minor vandals, violent and into a spot of experimental petty theft.

The compulsory “sex abuse” cases were reported – some sort of kiddie porn fights back – I struggle to conceive of what this stuff involves, but that too is likely a function age – my old age.

A first lesson in statistics is to think about context and materiality. Much of the statistics discipline is concerned with figuring out what is and is not “significant”. The easiest tool for doing this is to apply the brain in thinking about the proportionality of what is involved.

Journalists should try it sometime.

No comments:

Post a Comment