Testing and assessment in English for Academic Purposes (EAP) contexts has
traditionally been carried out on the basis of a needs analysis of learners
or a content analysis of courses. This is not surprising, given the domina
nce of needs analysis models in EAP, and a focus in test design that values
adequacy of sampling as a major criterion in assessing the validity of an
assessment procedure. This article will reassess this approach to the devel
opment and validation of EAP tests on the basis of the theoretical model of
Messick (1989) and recent research into content specificity, arguing that
using content validity as a major criterion in test design and evaluation h
as been mistaken.