Are C-tests valid measures for L2 vocabulary research?

Abstract

Second language (L2) researchers (Singleton and Little, 1991) have sug gested that C-tests, developed as norm-referenced measures for proficiency and placement testing (Klein-Braley, 1985), can be used in L2 vocabulary research. This article illustrates how researchers can bring to bear essentials of measurement theory on L2 research by weighing validity justifications pertaining to use of the C-test method for vocabulary assessment in L2 research. Validity is defined using the predominant framework from current measurement theory (Messick, 1989) and its relevance for L2 research is explained. The cornerstone of the definition is construct validity, which requires a definition of the construct to be measured - interlanguage vocabulary (i.e., vocabulary ability). A theoretical definition of vocabulary ability is presented and used to consider justifications for and against interpreting C-test performance as indicative of vocabulary ability. On the basis of evidence concerning construct validity and utility as well as the consequences of interpretations, the potentials and limitations of the C-test method for L2 vocabulary research are identified.