ABSTRACT
The reasons for using tests are bound up with the nature of the information they yield; and this depends entirely on the features built into a test. One justification which is frequently given is that tests are ‘objective’, and so free of the biases that can affect other forms of assessment. On its own, however, this reason is not sufficient. Also, as we shall see, objectivity is a matter of some debate. What really matters is whether or not a test produces good quality information. In beginning to think about quality we need to ask fundamental questions, such as: who is being assessed; who are the assessors; what do we know about the attributes assessed and how are they defined for testing?
