Knowledgebase: Key Publications
'A-level Results in Comprehensive Schools: the COMBSE project, Year 1' by Carol Taylor Fitz-Gibbon
Posted by Paul Skinner on 29 January 2013 03:46 PM

The paper A-level Results in Comprehensive Schools: the COMBSE project, Year 1 was written by Carol Taylor Fitz-Gibbon in 1985 while at the School of Education, University of Newcastle-upon-Tyne.

The paper analysed A-level English and Mathematics results from ten comprehensive schools in the north-east of England.  The Confidential, Measurement-Based, Self-Evaluation (COMBSE) project enabled participating schools to compare their A-level grades, taking into account a wide range of intake variables.

  • The teacher's rating of the pupil's ability was shown to correlate most highly with A-level grade.
  • Average examination grade at the end of compulsory schooling was shown to be the best single, objective predictor of A-levels .
  • A negative correlation was found between time spent on homework and grades achieved.
  • For English, smaller classes were associated with lower residual gains (value-added).  This supported findings of a report from the Department of Education and Science (DES 1981) that schools with small sixth forms had been obtaining poorer results.
  • On the basis of average O-level grades the English candidates were 0.3 of a grade less able than the Mathematics candidates.

As well as laying the foundation for the continuingly successful ALIS Project and other monitoring systems from CEM, the paper questioned the reasonableness of the required 30% A-level failure rate and relative difficulty of Mathematics to English.

Key words: A-level; average O-level score; Confidential, Measurement-Based, Self-Evaluation (COMBSE); Advanced-Level Information System (ALIS); Centre for Evaluation & Monitoring (CEM); Value-Added; Homework; Failure rate; Relative difficulty; Least Significant Differences; Scheffe


"A-levels ... may be the most decisive examination students ever take."

"... the mean O-level grade provided a better predictor than single subject grades or measures reflecting quantity."

"... pupils from more educated and professional homes did slightly better than expected in English. In Mathematics home background had no detectable effect ..."

"Pupils learn from each other as well as from the teacher."

"For comprehensive schools, which now educate the vast majority of children, the important task is to monitor results and interpret them with reference to intake characteristics in order to make sure all pupils are getting a fair chance to succeed in a sensible system of examinations."

"Schools will be the final interpreters of their own particular set of results."

"The school's own observations and interpretations, discussed with researchers, are an invaluable help in the generation of hypotheses and therefore in guiding the development of the instruments year by year."

"... the important question will be whether the same classes are doing relatively well and relatively poorly in subsequent years. Without such stability in the results they are hardly worth examining."

"... important statements about individual schools will only be possible after several years of data have been examined, and then only if the findings from year to year show consistency."


DES (1981). Statistical Bulletin 8/81. The relationship between size of sixth form, teaching group size and “A” level performance in maintained schools. London: H.M.S.O.

Comments (0)