## Monday, November 14, 2011

### Predicting Success at the Eleven Plus

Some parents often ask the question: “With these scores, will my child pass the eleven plus?”

Other parents may ask: “As a result of these scores - what are the chances of my child passing the eleven plus?”

The parents are asking for some form of prediction. A predicted score is a kind of average score. When eleven plus results place a number of bright eleven year old children into the care of the grammar schools, parents are predicting that the school will draw the best out of their children – and that the Year 7 pupils will go on to have stellar GCSE and A Level results. Some parents may think: “On an average children who go to grammar school usually do well academically. If my child goes to grammar I can expect great things.”

If a parent went up to a complete stranger in the holiday resort of Disneyland Paris and asked: “Will my child pass the eleven plus?” The respondent could answer in a number of ways. “What is the eleven plus?” Who is your child – do I know him?” “Please – I do not speak good English – but I will try to help.”

Of course you may be lucky enough to meet the educational psychologist who last year placed your child in the top 2% of the population with an I.Q. of 162 – who would be able to reassure you of your child’s chances!

In any average score it is likely that there will be some form of range of results. It is this range that makes the powers of prediction so uncertain. A parent may argue that scores around 80% on eleven plus papers is supposed to be able to suggest that their child can pass the eleven plus. If the child has done two papers – and both of the results are over eighty – then a score of 80 might be enough. If the child, however, worked through a range of papers from different sources – and still achieved an average of over 80% - then this could, possibly, be more reliable.

A statistician has the ability to compare an Expected Result with an Actual Result. The difference between the ER (Expected Result) and the AR (Actual Result) may be small in some cases – but remarkably large in others. A parent, for example, may forget that they `help’ their child with odd questions while the child is doing the test at home. (The act of forgetting is called a Selective Memory.)

It the difference between that predicted scores and the actual scores is remarkably small – and possibly of no real consequence – then it is very unlikely that an error in selection has taken place. With a large gap between scores – then errors could creep in.

Suppose a Local Authority wanted 20% of the children who took the eleven plus examination to pass. This would mean, hopefully, that twenty children out of every hundred would earn a grammar school place. If, however, the authority picked – at random twenty children out of every hundred children – then some bright children would, possibly, win a place, some border line children could also be successful and other children (who may find grammar schoolwork too hard) would also have a chance.

We could argue, for example, that 30% of children in a Local Authority have the ability to cope with a grammar school education. If the Local Authority sets a limit of wanting just the top twenty per cent in the grammar schools then some very bright children may miss out.

"Yes but, will my child pass?"

"It depends."