Value Added and Grammar Schools
by Mark Tweedle, Head of Heckmondwike Grammar School

What is value added?

This is a measure of the difference between the actual and expected progress of a pupil or cohort in a Key Stage (KS).*  Positive value added means that pupils have exceeded expectations, negative value added means that pupils have not achieved the expected levels. The expected performance is based on prior attainment and is an extrapolation based on pupils maintaining their overall level within a cohort. The value added index published in the tables is (100 + calculated value added index).  A value added score of 100 represents zero value added, ie a pupil or cohort has achieved in line with expectations.

There are separate value added calculations performed between KS2 and KS3 and between KS3 and KS4.  The expectations for KS4 are revised in terms of the outcomes from KS3. The DfES is currently conducting a pilot to measure the value added between KS2 and KS4 which omits the use of KS3 data. [*Note for parents: KS1 covers ages 5-7,  KS2 ages 7-11, KS3 ages 11-14 and KS4 ages 14-16.  Pupils do national tests at the end of  KS1, KS2 and KS3, and GCSEs are usually done at the end of KS4. The most recent value added tables for KS2 to KS3 (covering ages 11-14) were published in December 2003, and those for KS3 to KS4 (covering ages 14-16) were published in January 2004.]

Why do grammar schools do so well in value added terms between KS2 and KS3?

Given that pupils entering selective schools have in general achieved highly in KS2 tests the expectations for their performance at the end of KS3 are high. On this basis one would expect that it would  be difficult to add significant value.  Indeed, simply maintaining the good progress achieved so far would represent a good achievement. In fact grammar schools demonstrate very good value added between KS2 and KS3, dominating the recent tables in terms of both attainment and progression.  This is because pupils in the selective schools achieve well beyond the normal expectation at the end of KS3 and  the grading of the KS3 tests allowed able pupils to demonstrate their very good progress.  Pupils achieving level 5 at the end of KS2 are in the top 20% of the national cohort - those who go on to achieve level 7 at the end of KS3 (as many pupils in selective schools do) have progressed to the top 10% of the cohort.  In mathematics progress can be more dramatic because pupils are still able to access level 8 and those that do so are in the top 3% of the cohort.  In selective schools well over two thirds of pupils achieve a level 7 or better, and that figure rises to over 90% in the highest achieving schools.  In value added terms,  because so many pupils in selective schools who achieved in the top 20% at the end of KS2 have moved into the top 10% or better, progress is significantly above expectation and there is strong value added. 

How did this year's KS3 added value achieved by selective schools compare with last year?

In last year's tables the value added performance of the selective schools was even better than this year  because the extension papers for English and Science were still available. These papers, which were designed for the most able, allowed pupils access to level 8 and EP (exceptional performance). Students achieving level 8 in either science or English were demonstrating a level of performance which placed them in the top 1% of the cohort.  Sadly, the uptake of the extension papers on a national scale was low and the extension papers were withdrawn this year.  The selective schools did however make good use of the extension papers to stretch their ablest pupils and this was reflected in last year's KS3 value added data.  Allowing pupils to demonstrate a level of performance which moved them from the top 20% of the cohort to the top 1% gave rise to very significant value added.

The added value data for selective schools seems less good at KS4.  Why is this?

This is because the methodology used to calculate added value effectively prevents the most able students in selective schools from adding any value at GCSE.   The expected performance of each pupil in KS4 is recalculated on the basis of their achievement in KS3.  In 2001 many pupils in selective schools achieved two or more level 8 grades in the KS3 tests, placing them in the top 1% of the cohort.  For those pupils to maintain this standard they would have to obtain GCSE grades in the top 1% of the cohort  ie A* grades.    Because the A* grade represents a ceiling of GCSE achievement,  a significant proportion of pupils in selective schools are not able to add value, the best they can do is to break even.  The methodology used further disadvantages selective schools by capping GCSE achievement to  8 GCSEs and discounting AS and A2 grades taken early.

Are there other problems with the value added methodology?

The performance of pupils in three tests in mathematics, English and Science taken at the end of KS2 is used to predict the performance of pupils in similar tests in the same subjects at the end of KS 3. The methodology is valid because the subjects and assessment techniques are the same. This is not true of KS4 where the three KS3 tests are being used to predict performance in a range of subjects assessed by a variety of methods. The correlation between KS3 and GCSE performance is less strong particularly in creative subjects or where coursework is the dominant assessment method. 

Are Grammar Schools over reacting to a set of apparently adverse data?

No, the opponents of selection are using this flawed data to discredit grammar schools.  Teachers and Governors in selective schools have no argument with league tables and statistics but the methodology used must be fair and transparent.   Selective schools consistently attain results  which place them in the top 5% of schools. Ofsted say they are well managed and led with high standards of teaching and learning. Value added tables say they help their pupils make outstanding progress in the first three years of secondary education. Why should schools which are noted for their consistently high standards make less progress with their pupils in Key Stage 4?   What possible explanation could there be for this surprising conclusion?   Is this a function of the schools or the statistical methods used to calculate the added value?  

Supporting data

The 2003 GCSE cohort entered secondary education in 1998.  The achievements of this cohort in the KS2 tests is shown in table1, below.

Table 1 Data showing the percentage of the cohort achieving each level in the KS2 national tests in 1998

This cohort took the KS3 tests in 2001 and their achievements are shown in table 2. 

Table 2 Data showing the percentage of the cohort achieving each level in the KS3 national tests in 2001

The level 8 grade was only available in Science and English if pupils were entered for an extension paper.  Nationally, the uptake for the extension papers was very low (they were abandoned in 2003) but much higher in the grammar schools. Almost two thirds of  the levels achieved by grammar school students in the KS3 national tests were  7 or above.  This figure rose to 90% in the highest achieving grammar schools.  The implication is that a significant proportion of the level 8 grades awarded in 2001 went to pupils in selective schools.

Pupils awarded level 8 in the 2001 KS3 tests were achieving in the top 1-2% of the population. The value-added methodology used by the DfES sets KS4 expectations for these pupils based on maintaining this position in the national cohort. To perform at this level requires pupils to achieve A* grades at GCSE. If they achieve the A* grade they will simply maintain the progress already made (ie add zero added value).  Because the A* grade is the ceiling of GCSE achievement it is not possible for these students to add further value in KS4.  The best that can be achieved in added value terms is zero.   The methodology used by the DfES penalises selective schools because they have a high proportion of able pupils who are unable to add value in Key Stage 4 no matter how well they do at GCSE. This is further compounded by the DfES decision to cap GCSE achievement at 8 subjects and to discount the results of pupils who take AS and A level examinations in KS4.

 Case Studies

Pupils A and B make the same progress in absolute terms (two levels ) between KS2 and KS3,  but the methodology used assigns considerably more added value to the lower attaining pupil.  In KS4,  both pupils again make similar progress in terms of the level of grade in each subject but pupil B achieves greater breadth.  The methodology discounts four of pupil B's twelve A grades and results in negative added value.  The lower attaining pupil, adds value, whereas the higher attaining pupil subtracts value.   A grades or better are typically achieved by the top 10% of the cohort.  Despite the fact that pupil B has progressed from the top 20% of the cohort in KS2 to the top 10% of the cohort in KS4,  the overall value added is zero (+2 -2).  The lower attaining pupil has made the same progress in relative terms but the methodology assigns a value added of +5 (+4 +1) over the two Key Stages. 

Pupils C and D demonstrate the very significant impact of  level 8 on KS4 value added.  In schools where a significant proportion of pupils achieved level 8,   able students who achieved full sets of A grades and better were subtracting value.

In the example below pupils E and F have the same starting and finishing point but pupil F adds substantial value between  KS3 and KS4 whereas pupil E subtracts value because of the difference in KS3 performance.  Interestingly,  if the overall value added in the two Key Stages is calculated  (by adding the two component added values together) the DfES methodology assigns more added value to the lower attaining pupil at KS3,  although logically there should be no difference.