Describes an alternative approach to assessing content validity and content equivalence in terms of item-content structures and content area constructs. Results from applying structural equation modeling to item-response data from two Regents College examinations to verify content constructs suggest the different degrees of inconsistency and bias of content experts in assigning items to their corresponding content areas. (SLD)