Extracted from 'Good Assessment Practice'.
Lewis
Elton, University College London
It may be obvious that what should be assessed in an educational programme is the extent to which its objectives are achieved, but in practice this is often not the case.
Different methods suit different candidates better or worse, so that a mix is fairer than a single method.
Quite generally, examiners might consider whether to give more choice to students, since choice - even between the frying pan and the fire - tends to motivate.
Coursework assessment is particularly suitable for assessing process objectives, and essential for assessing true creativity and genuine problem solving abilities, neither of which can normally be assessed under the stresses and time pressures inevitable in formal examinations.
Reliability. There are two kinds of reliability, which a measuring instrument must satisfy.
Validity. An assessment is valid, if it assesses what it is intended to assess, which is usually specified in terms of the learning objectives which are to be assessed. It is generally agreed that learning objectives must not be specified either too tightly or too loosely, if the learning is to be validly assessed,
The relationship of reliability and validity. There has to be a trade off between reliability and validity.
The 'backwash' effect. Students' learning is guided by the assessment to come and the objectives being assessed become the students' learning objectives.
Connoisseurship. Connoisseurs are persons who, through training and experience, can make expert and reliable judgments in their specialist fields, which is not a bad definition of a good external examiner.
Constructing essay questionsParticular care must be taken with essay questions, which usually contain an operational word, such as 'discuss', 'compare', etc. Possible operational words are:
© Lewis Elton 2002
![]()