Rosemarie Frost points out in her letter that exams which are too hard demoralise students (TES, October 1). A more serious problem is that excessively difficult tests produce unreliable results. The depression of scores inevitably results in grade boundaries which are very close together and, since all tests have some error of measurement, this causes many students to be mis-graded. The subsequent transformation of scores to a longer scale (UMS) does not solve this problem - it simply hides it.
George Bethell Director, Anglia Assessment Valley Road Battisford, Suffolk