This paper replicates Hemchua and Schmitt’s (2006) study into types and frequency of lexical errors in Thai university students’ compositions. To investigate the usability, reliability and validity of their framework, 20 Greek learners’ compositions were analysed, following the original methodology. Results concerning the number, distribution and frequency of lexical errors were remarkably similar; approximately one third of all errors were formal, two thirds were semantic and less than 13% were attributable to transfer. Four of the five most common sub-categories of error in the replication were also found in the most common five sub-categories in the original study, suggesting that the framework, when applied to a different context and nationality, produces similar results and may reveal common problems between different English learners with different first languages. Difficulties in error identification and categorisation are discussed in detail, and suggestions for development of an improved framework for analysing lexical error are made.
This work by European American Journals is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License