Abstract

The traditional, ‘pre-unified’ conception of validity at least appeared to give clear guidance on validation: if you are validating for this reason (e.g. to predict future job performance) then collect this kind of evidence (e.g. criterion-related validity evidence). Many commentators have argued that the ‘unification’ of validity theory – through the notion of construct validity – has made it far harder to understand what validation ought to comprise. Reports of a disjunction between modern validity theory and present day validation practice are increasingly common; and the ways in which people describe and characterise validity, even within the literature, are frequently equally out-of-kilter. What are we to make of this disconnection? Is it possible that modern validity theory is just too abstract, confusing or unclear to scaffold effective practice? Participants will be encouraged to examine the extent to which this may or may not be true.