Abstract
The Angoff procedure for setting cut‐scores for criterion‐referenced tests has generated considerable interest in the UK. The method offers a means of harnessing teachers’ experience of children's performance at different National Curriculum levels to the problem of setting level boundaries in tests with a continuum of marks. The research reported here examines, from several perspectives, the appropriateness of a cut‐score derived from the Angoff procedure for a reading test for 7‐year‐old children. This shows the recommended cut‐score to have been too low, and it appears that this could have been due to teachers’ lack of experience with the type of test concerned. Suggestions are made as to how standard setting might draw on a range of information to produce an appropriate and rationally defensible cut‐score.