Friday, 4 October 2013

Possible increased risk of serious abuse among home educated children in the United Kingdom

We looked yesterday at some cases of the murder of home educated children in the United States. Commenting on this, one person made the assertion that; ‘I can't see that it is more likely to afflict home educated children than others.’  This of course is a common claim by home educators in this country; that home educated  children are at no greater risk of abuse than schoolchildren.  If you repeat something often enough and loudly enough, then after a time, people sometimes begin to believe that it is true, simply by virtue of their having heard it said so frequently.  Perhaps it is time to examine this particular  statement and see what we can make of it.

One of the problems with calculating rates of abuse, or educational attainment or anything else for that matter, among home educated children is that we don’t really know how many there are. We know that 20,000 home educated children are known to local authorities in England and Wales and that there are also some children unknown to local authorities. Let’s assume that for every  home educated child known to local authorities, there is another, about whom they do not know. This would give us approximately 40,000 children. Expressing this as a percentage of children aged between 5 and 16, would suggest that something like 0.5% of children in this country are home educated. This is around one in two hundred. So if we take a random sample of two hundred children, we might expect to find one home educated child among them. These are of course very rough figures.

Suppose now,  that we take a group of 36 children aged between 5 and 16. If we have drawn them at random, then there is a slim chance of finding a home educated child among those 36 children. The odds are 6 to 1 against the group containing a single child who is home educated. Still, even if you found one or two, that wouldn’t really signify. Coincidence and chance play strange tricks sometimes. If however you found that  8 or 9 of that group of 36, 25% of the total, were home educated, then you would know that this was an atypical group, drawn from a section of children  which was far more likely to be home educated. The proportion of home educated children would, after all, be around 50 times as high as in the general population. 

Let us now look at such a group of 36 children. Here are the statistics for every serious case review evaluated  by Ofsted from April 1st 2010 to September 30th 2010; a six month period. It is important to realise that these are the statistics for every SCR, there has been no cherry picking;

These cases involve 93 children in total. We are interested only in those children of compulsory ’school’ age, that is to say between the ages of 5 and 16. We find that there are 36 such children.  If we look at sections 34 and 35, we see  2 cases which  cover 7 home educated children in 2 SCRs. We also see that there was another case which involved home educated children, although we are not told how many children were involved in that one. If we assume that it was 1 or 2, then that would give us 8 or 9 home educated children out of the total of 36 who were covered by the serious case reviews in that six months of 2010. What we are seeing here is a proportion of home educated children in this special group of 36 children who were involved in serious case reviews that is 50 times the rate in the general population of children aged between 5 and 16. 

I find statistics such as these quite intriguing. Although the samples are small, I think it fair to say that we can see many times  more home educated children than we would expect, all else being equal. Of children aged between 5 and 16 involved in serious case reviews over that six month period, 25% were educated at home; 50 times the percentage one would expect in a random sample. It is statistics  like this that make some professionals uneasy. Of course, the figures we have used are very rough and ready;  the true proportion of home educated children identified in cases of serious abuse in a random six month period might be  a little less or a little more than 50 times as high as in ordinary society.  We are not talking here though about some minor variation that only a trained statistician could identify; home educated children are hugely over-represented in this sample.  I will be busy for a few days, but will be looking again at this question soon.


  1. Is this trend continued through other periods, or is it a one off, Simon?

    Because my perception of the SCR's that are making the news is that there are common features that do not involve home education, but do involve professionals who, for whatever reason, fail to act on concerns when dealing with particular types of familes.

    It would, of course, be totally wrong to say that all single mothers living in Birmingham were likely to be child abusers so needed special supervision. It would be equally wrong to say that all home educators needed special supervision based on a single sample.


    The 'hidden' child scenario doesn't seem to be happening with respect to HE. Both Khyra Ishaq and that poor girl whose mother decided artificial insemination was a good idea were known to the authorities.

  2. 'Is this trend continued through other periods, or is it a one off, Simon? '

    That's an interesting point. SCRs don't always mention the educational setting, especially if it isn't relevant to the abuse. The evaluation in 2010 probably make a point of identifying which cases had home education as a feature, because of the Badman review the year before. I have certainly seen serious case reviews where I know that one of the kids was home educated, but it hasn't been mentioned.

    'It would be equally wrong to say that all home educators needed special supervision based on a single sample.'

    True, but then that wasn't really what I was suggesting. As I'm sure we both know, the claim is very often made by home educators that there is no evidence to suggest that home educated children are any more at risk of abuse than those sent to school. I was just pointing out that there is evidence that this might be so, by showing that of a random sample of serious abuse cases, a quarter involved home educated children; roughly fifty times as many as one would expect. There's more evidence and I hope to explore this further in later posts.

  3. Been away - so not avoiding issue - will plough through the stats