Not Saussure

June 3, 2007

Stats and moral panic

Filed under: Panic, Spin, Stats — notsaussure @ 4:05 pm

This, I’m afraid, repeats a discussion we’ve been having at Casa Worstall, but I thought it worth rehearsing here as an example of how statistics get twisted. I’m not trying to minimise the problem at all — we deal with it in the courts with depressing frequency — but it seems to me that there’s no point in exaggerating a problem by misreading what a report actually says.

Tim quotes Minette Marin in the Sunday Times:

Incredible though it may seem, there are hundreds of thousands of paedophiles living among us, perhaps next door or on the next floor. That, at least, is the estimate of the National Society for the Prevention of Cruelty to Children and nobody has seriously questioned its research. According to the NSPCC, 16% of all women and 7% of all men interviewed said they had been physically sexually abused before they were 12. That would amount to one in nine children.

These alarming figures come from Cawson et al., 2000, Child Maltreatment in the UK: A Study of the Prevalence of Child Abuse and Neglect, NSPCC. This isn’t available on-line, but an executive summary is, so let’s see what they have to say about Sexual abuse:

Numbers of respondents recording sexual activity with relatives which were against their wishes or with a person 5 or more years older, were very small: 3% reported touching or fondling and the same proportion had witnessed relatives exposing themselves. The other categories of oral/penetrative acts or attempts, and voyeurism/pornography were reported by 1%. Much larger numbers had experienced sexual acts by non relatives, predominantly by people known to them and by age peers: boy or girlfriends, friends of brothers or sisters, fellow pupils or students formed most of those involved.

The money quote, really, is right at the end of this section of the summary:

When respondents with experience of sexual acts as defined above were asked if they considered their treatment to have been sexual abuse, 6% of the total sample considered they had been sexually abused. There was overlap with the research defined abuse for those abused by relatives but much less so for others.

So that’s 6% — which is one in 16.7 rather than one in 9 — who actually ‘said they had been physically abused’. Still alarmingly high, perhaps, but not 1 in 9.

Furthermore, that’s 6% of the whole survey group — children up to the age of 16. I’m not completely certain, but I think what Minette Marin’s done is misunderstood the figures given in the section on ‘prevalence of abuse’, which is admittedly rather confusingly written, since it’s not at all clear when the figures refer to abuse in the sense of sexual activity between someone under 12 and someone 5 or more years older, whether ‘consensual’ or not, and when they refer to non-consensual sexual activity with ‘age peers: boy or girlfriends, friends of brothers or sisters, fellow pupils or students,’ and I think that to get 1 in 9 you have to do some double counting.

As I say, though, it’s clear from the summary overall that by far the greatest amount of ‘abuse’ as reported here comprises non-consensual sexual activity and sexual assaults by children of roughly the same age; and while there are all sorts of terms one might justifiably apply to a 14-year-old boy who sticks his hand down his 13-year-old girlfriend’s knickers without her permission, or grabs hold of a class-mate’s breasts, ‘paedophile’ isn’t one of them.

March 18, 2007

Skunk and schizophrenia

Filed under: Medicine, Mental Health, Stats — notsaussure @ 8:23 pm

Among the Sindy’s reasons for no longer wanting cannabis legalised (though they’re still in favour of decriminalisation):

25,000 schizophrenics could have avoided the illness if they had not used cannabis

Anyone got any ideas about how that figure’s arrived at? How can they say with any certainty that it’s nearer 25,000 rather than nearer 12,500 or 50,000?

I’m sure there is a good reason for using that figure, but any calculation, I’d have thought, must involve an awful lot of rather speculative assumptions about schizophrenia, what causes it and what triggers its onset.

Technorati Tags: , ,

Blog at