Many years ago an advert encouraged women to buy a new style measuring tape:
LADIES BE AS
AS YOU LIKE
PATENT EASY STRETCH
This is a bit like believing that the eleven plus examination will be fair to all children. Some children will no doubt benefit from an examination based on verbal reasoning questions – but other children may need to be able to demonstrate other skills.
One way to achieve test consistency is by offering the children a test – and then retesting them on a similar test a little bit later on. There would be no proof of inconsistency if different results were obtained – but the greater the divergence between the two sets of marks would point to some degree of inconsistency.
If there was perfect correlation between the two sets of test marks then it could be argued that the examination was valid. One county publishes the results of the top 180 children – so if the same 180 children achieved the same positions on the retest then the test could be considered to be reliable and valid.
This information would only be useful to parents and children if they felt that the test questions were fair and reasonable.
We need an example.
If an eleven plus mathematics examination was set to include lowest terms in fractions then all the children would ideally need to be able to do lowest terms. Does the examiner make all the questions easy so that all children can do the questions – or are the questions made hard so that only the top twenty five percent can provide the correct answers?
The key point here is – what happens to the children who are in the top 26 percent and not the top 25 percent? Should they be denied a grammar school place?