Validity of assessment

In the case of the standards model, by way of contrast, validity resides in the interpretations and uses to which test scores are put (Messick 1989), that is, in the test's alignment with the total teaching context. For example, if sitting an exam results in students rote-learning model answers, then that is a consequence that invalidates the test. An aligned, or properly criterion-referenced assessment task is valid, a non-aligned one is invalid.

In measurement model: External: how well the test correlates with outside performances Selecting students.

In Standards Model: Internal: how well scores relate to the ILOs and to the L\rget performance domain

Are they assessing what they should be assessing – are they valid?

In the measurement model, the test needs to be validated against some external criterion to show that the trait being measured behaves as it should if it were being measured accurately.

In the case of the standards model, by way of contrast, validity resides in the interpretations and uses to which test scores are put (Messick 1989), that is, in the test’s alignment with the total teaching context.

Source: Biggs, J.B., & Tang, C. (2007).