ABSTRACT
A fundamental notion in test validity is that the scores of a test should not be significantly influenced by factors other than the skills and knowledge that the test is intended to measure (Messick, 1989). This notion is critical to examining language issues in testing. Since language is the primary vehicle for administering tests, academic achievement measures are largely influenced by language variables, even when the constructs of interest are not directly related to language. That tests inevitably assess language skill in addition to the target knowledge domain (AERA/APA/NCME, 1999; Kiplinger et al., 2000; LaCelle-Peterson and Rivera, 1994) appears not to be an overstatement.
