Abstract
In 1997 a system was introduced to moderate variability in teacher marking standards in a 'Communications for Professional Engineers' subject. The paper describes implementation of the moderation procedures and of methods used to determine the reliability of teacher assessments. Results from the study indicate that moderation had improved the reliability of teacher assessments, but resulted in very minor changes in the marks for most students. Outcomes of the study are discussed in relation to the importance of assuring the fairness and consistency of marking in situations where assessments were perceived to be highly subjective.