From 1998, primary schools will have to assess reception pupils according to new guidelines announced this week. The School Curriculum and Assessment Authority drew up three model "baseline assessment" schemes, which went out for national consultation last term. At the same time, the National Foundation for Educational Research was commissioned to carry out independent field trials of these models. The aim was to find out two things: how far did each model provide a fair assessment, despite differences in the children's ages and backgrounds? And how effective were the models in providing teachers with useful information, a manageable method, and a reliable baseline measure?
Our research involved 420 teachers and 2,647 children in 307 schools, and,using the checklist that featured in two of the prototype schemes, shows what proportion of the children came to school with specific early literacy and numeracy skills, such as recognising letters and counting objects accurately. This checklist proved to be pitched slightly too high,largely because so many of the children starting school are only four, while the list was based on expectations of children turning five. The average score was eight out of 20.
Scheme 1 was based closely on the Desirable Outcomes booklet published by SCAA and the Department for Education and Employment a year ago. The booklet sets out the standards expected of children when they start school,assuming they are already five and have received some pre-school education. It identifies six areas of learning: personal and social development, language and literacy, mathematics, knowledge and understanding of the world, physical development, and creative development.
For each category, teachers were asked to decide whether the child was at grade "A", "B" or "C", using performance descriptions.
Teachers expressed two main concerns. Most said they did not find it easy to apply these criteria in practice. They were not sure what to do when a child's performance across the different areas was not consistent. Around two-thirds said they did not find it helpful.
This complexity also meant that the scheme did not give useful information for planning learning activities. Nearly 80 per cent found it "of limited use".
Our second main finding was that the middle grade proved, on average, to be too difficult for children in their first half-term of reception. A grade of C was most common. This was mainly because these children did not meet the profile for the Desirable Outcomes standard. Their average age was only four years and nine months, and some of them had not had any kind of pre-school education.
As might be expected, the children who had no pre-school education did less well than those who had. The differences between types of pre-school provision (nursery, playgroup, day nursery) were not significant. Similarly, children learning English as an additional language - of whom a separate study was done - did less well. Older children in the year group did better than younger children, as one might expect, and girls did better than boys.
SCAA's other two schemes both used the 20-point checklist of reading, writing and number skills. Scheme 2 consisted of the checklist alone, while with Scheme 3 it was supplemented by a descriptive record.
The checklist covered only three areas of the curriculum and used much simpler criteria than the first model. These criteria proved much easier to use than the performance descriptions, with nearly two-thirds of teachers finding them helpful.
However, teachers reported other difficulties. Many saw the checklist as too narrow. In particular, they wanted to include personal and social development, speaking and listening in their assessments.
But the main criticism concerned the "yes-no" nature of the checklist. If a child did not get a tick, no indication was given of what they actually could do. Some teachers were also worried that they were labelling children as failures by simply recording "no". This was a particular problem for teachers of children learning English as an additional language, who pointed out that the physical surroundings and social expectations may be unfamiliar as well as the language. These children might achieve no score at all on the checklist, thus giving no indication of their level of development on starting school.
Again, the checklist seemed to be pitched too high for children who were in their first few weeks of school.
In Scheme 3, the simple checklist was supplemented by a descriptive record. At the top of each page was a heading related to one of the six areas of the Desirable Outcomes. Apart from this, the record was blank, leaving teachers to write whatever they wished.
These records proved to have a good deal of breadth and flexibility. They allowed teachers to record even the smallest steps of attainment, and to build up a record of observations of each pupil. Scheme 3 attracted strong support from some teachers, who valued the information on individuals, and regarded it as a fair way of assessing all children.
It was, however, very time-consuming. Almost half of the teachers trialling Scheme 3 reported that it had taken them more than five hours to record the assessments, and that was just for the six children included in the trial. Many teachers said that, despite its advantages, the scheme would not be practicable for a whole class of children.
Although none of the schemes proved entirely suitable, the opportunity to perform a full range of analyses on a large national sample has given some clear pointers for future work.
To be easy to use, the criteria should not be too complex. Baseline assessment should be broader in scope than the checklist. The assessment criteria should be graded so that the attainments of the full range of children can be recognised and teachers have information to plan learning. They should include points below the standard of Desirable Outcomes, as well as one point above.
The materials that we are currently developing for SCAA, for optional use nationally, will build on these findings by using straightforward criteria arranged into four-point scales covering the whole range of attainment.
Dr Marian Sainsbury is a senior research officer with the NFER.
The full report, Trials of Baseline Assessment Schemes, by Marian Sainsbury, Louise Caspall and Catherine Kirkup, with Ian Schagen and Chris Whetton, is available from SCAA, Newcombe House, 45 Notting Hill Gate, London W11 3JB