ABSTRACT
In this chapter, the authors discusses the problems involved in moderation and ways in which these might be solved. The essence of the problem can perhaps be most readily appreciated by envisaging the outcome of a series of school-based examinations. The range estimate is a measure of spread which is extensively used in certain statistical procedures applied to industrial processes but one which has never before been introduced into educational measurement. The authors examine the possibility of conducting school-based examinations from the results of which valid grades could be determined without the need for subsequent processes of moderation. Fortunately a somewhat simplified procedure has been evolved for the conduct of local moderation—that is the moderation, by a panel of teachers, of the marks or grades submitted by a small group of schools in the immediate vicinity.
