Abstract
Much of the research initiated by examining bodies in Australia has concerned the standardizing of marks and grades to achieve comparability when different methods of assessment are used to evaluate the same course objectives and between subjects whose candidatures are not homogeneous. The procedures employed raise theoretical issues, some of which (the interpretation of grades, meaning of comparability and use of moderator variables) are discussed in this paper. It is shown that although no numerical transformation can make measurements of different skills comparable, performances can be compared, in a relative sense, by reference to a standard population. Scaling, then, is seen as the process by which the grade relative to this population is estimated from the raw mark. Two procedures which employ moderator variables are examined and their characteristics noted.