High-stakes assessment in England and Singapore

Its annual "Autumn Package for Pupil Performance Information," provides tables with the percentage of students nationally (overall and by gender) reaching reading, writing, and mathematics benchmarks, provides a column in which to enter the school's results, and another to calculate the difference in percentage points (Department for Education and Skills, 2001). The department has also developed a "Five Stage Cycle for School Improvement" program to enable educators to review SATs data for evaluation, target setting, planning, and monitoring. Fundamental to the cycle are three questions: How well are we doing? How do we compare with similar schools? and What more should we aim to achieve? Within each of the five stages, there are roles for team and subject leaders and for classroom teachers, in general agreement with a hierarchical management structure.

Not surprisingly, the questions posed by the department focus on student SATs achievement and what can be done to raise that achievement. They do not ask questions about the suitability of the National Curriculum for the school's students, the validity and reliability of the assessment, the educational value of the assessment, the validity of comparing different classes or schools, and so on. The lack of questions that explore the implicit philosophical underpinnings, learning theories, and curriculum implications of the assessment severely limits the opportunities for teachers to participate as critical professionals, and offers no opportunity for those closest to the SATs to provide feedback to policy makers. Consequently, teachers are, in effect, treated as skilled technicians.

There is some evidence of a scapegoat effect in the government's handling of criticisms of its educational policies. Typically, the government defends its testing policies by deflecting blame for problems toward schools, teachers, or parents. For example, one of the checks used to establish that a high-stakes test is performing as expected is to compare the results it gives with results from another test that has significant conceptual overlap. When results from the reading ability and writing skill tests for 11-year-olds showed a marked gap, a spokesman for the National Primary Heads Association challenged the results, claiming that there were obvious errors in marking the writing test (BBC News, 2000b). However, the Office for Standards in Education (OFSTED), a government department responsible for inspection of schools and Local Education Authorities, investigated the issue and concluded that there was not a problem with the test, but that more attention should be given to training teachers to teach writing effectively. This conclusion was arrived at despite the fact that a review of the writing test papers in one school resulted in an overall 23% increase in the average test score.

Observing that rising test scores demonstrate little more than teachers' increasing abilities to teach to tests, the limited reliability of educational assessments, and the increased narrowing of the curriculum as teachers teach to the test, Wiliam (2001) noted that the "level of assessment illiteracy by successive governments is appalling" (p. 1).