THE Scottish Parliament has been debating 5-14 assessment and - judging by extracts published in The TES Scotland - MSPs have little understanding of the place of testing or of current practice.
Let’s hope that the Education Minister’s action group provides him with clearer advice than he receives from politicians. In its first 10 years national testing has made a valuable contribution to primary school assessment and self-evaluation but its flaws have long been obvious to teachers.
Tests are not as objective as they might be since they can be manipulated easily, well before the marking stage. With no embargo on the opening of test packs, a teacher has access to the questions long before the children do and it is possible to weigh one’s teaching towards the skills and knowledge required by the paper, while during the test a cough, raised eyebrow or carefully pointed finger can draw a child’s attention to any question which requires a revised answer. Keeping test papers secret will not achieve much in itself.
Present advice to test children when they are ready means that testing at any level, but particularly at level C, can happen throughout the year across a school with the contents of papers being known widely. Three testing periods per year, agreed within a school, would counter this difficulty, especially if different sets of tests are used. Supervision of the test by someone other than the class teacher would remove the temptation to help by surreptitious means.
The 5-14 curriculum levels are broad and it has been difficult to educate parents in their use as indicators of a child’s progress, especially when the two-year level C makes it appear as if there has been no progress during the first year after level B. Official statements are of little help and many schools have to rewrite attainment descriptions to help parents’ understanding. Annual progress steps and clear explanations of levels would be much more helpful.
Recent research confirms what many class teachers have discovered for themselves. Comparing a class set of 5-14 levels, say for reading, with standardised scores for the same class throws up many mismatches with the standardised scores indicating some children are assessed at levels that are too high or too low. This issue should have been addressed at the outset of national testing and must figure in any overhaul if tests are to gain real credibility.
However, it is not a simple matter of replacing national tests with standardised ones. They have problems, too, due to statistical limitations. Misinterpretation by teachers and parents is common and working in standard deviations would also leave us reporting a child’s achievement in broad bands just like a national test.
However, if the action group wishes to earn the gratitude of teachers, the task at the top of its list must be to improve the confusing criteria for the assessment of writing. Nothing is guaranteed to cause as much friction among teachers or between primary and secondary as this.
The minister also wishes to integrate the Assessment of Achievement Programme (AAP) testing regime with national testing. AAP results are derived from a 5 per cent sample of selected age groups to give a national picture of achievement every few years and, whatever the subject, generally tell us that P4 is fine, P7 poor and S2 pretty awful. If national testing becomes more reliable, there will be an annual picture of achievement in key areas and the AAP’s reputable bank of test items may be the place to start in the quest for consistency.
The flaws in national testing need to be ironed out but the action group must remember that testing is only one part of the assessment undertaken daily in each classroom and that teachers cannot cope with any “improvement” which uses more paper and more time than at present.
