Using frequency lists to assess L2 texts

• Main Author: Bell, H. M.
• Published: Swansea University 2003
• Subjects: 410
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.636075

Studies of lexical richness fall into three main domains: the analysis of classical literature, the study of first language acquisition, and the study of second language acquisition. In the last-names, a variety of measures of lexical richness have been proposed and used, but the measures are often seriously flawed in either design or interpretation. Recent studies have suggested that using text-analysis tools based on lexical frequency counts may be a useful contribution to the field. This thesis discusses the various methods of using frequency counts in this way, examining in details two particular measures: the Lexical Frequency Profile (Laufer and Nation, 1995) and Plex (Meara, 2001). My conclusions are that the use of lexical frequency data can provide a useful way of avoiding the problems associated with measures based n the type-token distinction and that the results can be presented in a straightforward and mathematical simple way. As a measure of lexical richness, Plex appears to be comparatively unproblematic, and a very useful development in this form of assessment. However, problems remain when this form of assessment is applied to the evaluation of the freely-produced language of learners. One reason for this is the difficulty of producing a frequency list which is appropriate to the situation of the learners.