ABSTRACT
The work undertaken in the PRISMS project was designed from the outset to be a study of small schools in their own right, not a comparison of small schools and large schools. In small schools, however, the fluctuations in the numbers of pupils entering school, not only in September but at other times in the year, are such that the definitions used by Lee and HMI are not entirely satisfactory. The survey confirmed that small schools were considerably more likely than a random sample of schools to have vertically grouped classes. Approximately 90 per cent of the PRISMS teachers who responded had more than one year group in their class, compared with under half of the non-PRISMS teachers. Although it was not part of the main thrust of the PRISMS project, people were able to devote a limited amount of time and resources to collecting data in large schools which were directly comparable with data collected in the small schools.
