There has been much interest in the statistics that the government is using in its campaign to link home educators with child abuse.
The essence of the story is that a survey of local education authorities has determined that 0.4% of home ed children are on the "At risk" register. This compares to a figure of 0.2% in the population as a whole. The 0.4% figure is described as varying greatly between different counties, suggesting to me that it is a measure that is prone to error.
The figures appear to be being used as a way of answering the question "Are HE kids more at risk of child abuse than those in schools", with "on the At Risk register" being used as a proxy for "At risk". It strikes me that these are not the same thing at all though. There are clearly very large numbers of children who are HE but are not known to the authorities and there will also be some who are at risk who are not known to the authorities either. Because of this, "On the At Risk register" would appear to be a very poor proxy for "At risk", at least as far as assessing HE is concerned.
The question is, how would you answer the question properly? With the survey as presented there must be a possibility that that the risk associated with HE is actually less than that for the population as a whole. But how would you calculate this probability?
This is a question for a stats blogger - I wonder if this man knows? I'll ask him.