Key Stage Four (Age 16)

Key Stage 4 data appears in 2 sets of tables

· A candidate level file containing summary achievement measures
· A qualification level file containing data about every qualification entered by a candidate

Data collection

Qualification data is collected from awarding organisations by the DfE Contractor RM Data Solutions who collect details of all qualifications entered by learners under the age of 19. Only qualifications approved for use under Section 96 of the Education Act are included.

These qualifications are then matched at learner level. Other sources of data are subsequently matched including measures of prior attainment (KS1, KS2, KS3) and School Census.

Summary measures of KS4 achievement are then calculated at candidate level, e.g. total points score and whether a candidate has achieved 5 A*-C (or equivalent) including English and mathematics. These measures are derived by applying business rules (calculations) at qualification level. Reference data is maintained to translate achievements in qualifications into performance measures. For example, a grade C in GCSE Short Course in History is equivalent to 0.5 GCSEs, contributes 10% towards the level 2 threshold and is worth 20 points. The contractor applies discounting rules to avoid double counting of qualifications, arising from resits or from qualifications taken en route to a larger qualification.

Further business rules are then used to determine

· Aggregate candidate level performance measures
· Enumeration flags that are used to determine which candidates to include in national, local authority and school level results.

The rules governing the above calculations change over time. For example, some international GCSEs were approved for use in 2010 (having not been approved previously) and hence contributed to performance measures.

Validity of measure

The matched KS4 data is provided to schools to check prior to publication of the DfE Performance Tables. All amendments are audited and either approved or rejected. NPD data includes approved amendments. Each year, a number of schools do not participate in the checking exercise (around 10% of maintained mainstream schools). From 2010, data on late awards and amendments from awarding bodies were included alongside amendments made by schools. Data that has been through the school checking process (whether the school participated or not) is given the ks4_prefix. Up to 2009, schools also checked School Census data. Hence up to 2009, differences may exist between the variables with _spr## suffix and those with ks4_ prefix, e.g. between fsm_eligible09 and ks4_fsm.

1.1 Candidate level file

The candidate file includes data drawn from

· The most recent school census (with _spr## suffix)
· Key Stage 4 (KS4_prefix)
· Key Stage 3 (KS3_prefix)
· Key Stage 2 (KS2_prefix)

In most cases, researchers will be solely interested in analysing the attainment and characteristics of learners at the end of key stage 4. In this case, select candidates where ENDKS=1 to remove other types of learner.

The file will include all learners who either

· Were on roll in NC Year 11 in a maintained school, CTC or Academy in January; or
· Were on roll in an independent school, had achieved at least one qualification and were identified by the school as a Year 11 student; or
· From 2010, were in local authority alternative provision (including Pupil Referral Units).

Some additional amendments are made to include pupils who were not on roll in Year 11 but who were found (from previous census records) to have completed their compulsory education in Year 10 (ks4_norflage=10).

Learners not included are those who were not present on any of the following administrative sources

· School Census
· Alternative Provision Census/ Pupil Referral Unit Census
· RM Bath Achievement Database

This will include:
· Those in education in other settings without qualifications (e.g. independent special schools)
· Those not in education

The absence of the former group means that a simple count of students in the file (where endks=1) will be lower than the total number of pupils at the end of KS4 reported in DfE Statistical First Releases. Similarly, it is not possible to replicate published statistics at national level from learner level data.

Generally, learners will only reach the end of key stage 4 once in their school history, in other words appear in one end of KS4 snapshot with endks=1. However, there may be one or two exceptions. Similarly, a learner should only appear at one school in a given year although there are known cases of duplicates where two schools could not come to agreement about which school should claim them for Performance Tables calculations.

Census and prior attainment data will generally be missing for those who were on the roll of an independent school (i.e. where KS4_MSCH=0).

1.2 Qualification level file

To identify qualifications included in a learner's summary end of KS4 measures, the exam level file first needs to be joined to the candidate file to identify relevant learners.

Qualifications that should be counted in a candidate's aggregate measures can be identified by (DISC3=0 and INCLUDE=1).

Cleaning the variable
Stata code
SPSS code
Checks needed

Description of values across cohorts

Pupil level measures of Total Key Stage 4 Attainment

There are two main types of measure of attainment at Key Stage 4 – point scores and threshold measures – and their measurement over time has changed. The former assign a score to combinations of qualification & grade; the latter are indicators that pupils have attained a certain level. There are different measures of each broad type of variable which depend on whether and how different qualifications are counted.

Point Scores

Scoring – Prior to 2003/04 point score measures were calculated using the ‘old style’ method of assigning 8 points to an A* grade in a full GCSE, 7 for an A, 6 for a B, etc down to 1 point for a G. Short course GCSEs are worth half the Full GCSE score. The ‘new style’ scoring system counts 58 for an A*, 52 for an A, 46 for a B, etc down to 16 for a G. More information about the change to the scoring system can be found at

Capping – Capped score originally was the total points score from a pupil’s best 8 GCSEs; under the ‘old style’ scoring this could be no greater than 64 and under the ‘new style’ scoring system it could be no greater than 464. In actuality, some (around 2% in 2010) of capped scores are over 464, due to the inclusion of early-taken AS-levels and that some equivalents qualifications are worth more than a GCSE A*.

Equivalent Qualifications – Points scores with ‘g’ in the title are just composed of results in GCSE and GNVQ exams, whereas an ‘e’ (equivalent) includes scores from equivalent (generally vocational) qualifications, which have become more prevalent since their introduction in 2004/05. Prior to their introduction almost all qualifications were GCSEs or GNVQs, so a ‘new style’ points measure can be constructed using the qualification information from those earlier years.

The variables that measure total attainment in 2010:
Total GCSE and equivalents new style point score.
2003/04 -
Pupil GCSE/GNVQ S96 Point Score (new points)
2003/04 - 2004/05
Capped GCSE and equivalents new style point score.
2003/04 -
Total GCSE/GNVQ new style point score.
2003/04 -
Pupil GCSE/GNVQ Point Score (new points)
2003/04 - 2004/05
Capped GCSE/GNVQ new style point score.
2003/04 - 2004/05
Total GCSE/GNVQ old style point score
2003/04 only
Total GCSE/GNVQ old style point score.
2001/02 -
Capped GCSE/GNVQ old style point score
2001/02 - 2003/04
Which measures are included changes over time as a result of the change from the ‘old style’ points measure to the ‘new style’ as the preferred score.


A number of different threshold measures are also included, concentrating on different margins and including different qualifications. As in the points score measures, some variables include equivalent qualifications and some don’t: FIVEAC/ FIVEAG include just GCSEs and GNVQs, whereas LEVEL2/ LEVEL1 include equivalents as well. The biggest change has been to using 5+ A*-C (or Level 2) including English and Maths, instead of the basic 5+ A*-C (or Level 2) as a headline measure in performance tables.
Achieved 5 or more GCSE/GNVQs at grades A*-C.
2001/02 -
Achieved Level 2 threshold (5 or more GCSE and equivalents at grades A*-C).
2003/04 -
Achieved 5 or more GCSE/GNVQs at grades A*-G.
2001/02 -
Achieved Level 1 threshold (5 or more GCSE and equivalents at grades A*-G).
2003/04 -
Achieved 1 A*-G at GCSE/GNVQ
2003/04 - 2004/05
2008/2009 -
Achieved at least 1 GCSE or equivalent at grade A*-G.
2003/04 -
Achieved any passes at all in GCSE or equivalents.
2004/05 -

5+A*C including English and Maths GCSE/GNVQ
2003/04 - 2004/05
Achieved 5 or more GCSE and equivalents at grades A*-C including GCSE English and Maths.
2003/04 -
Achieved 5 or more GCSEs (excluding equivalents) at A*-C including GCSE English and Maths
2009/10 -
Achieved 5 or more GCSE and equivalents at grades A*-C including GCSE and equivalents in English and Maths (Functional English and Maths).
2005/06 -
5+A*C including English, Maths Science GCSE/GNVQ
2003/04 - 2004/05
5+A*C including English, Maths Science GCSE/GNVQ S96
2003/04 - 2004/05
5+A*G including English and Maths GCSE/GNVQ
2001/02 - 2004/05
Achieved 5 or more GCSE and equivalents at grades A*-G including GCSE English and Maths.
2003/04 -
Achieved 5 or more GCSE and equivalents at grades A*-G including GCSE and equivalents in English and Maths (Functional English and Maths).
2005/06 -