Beginning and intermediate L2 writer's use of N-grams: An association measures study

Acommon approach to analyzing phraseological knowledge in first language (L1) and second language (L2) learners is to employ raw frequency data. Several studies have also analyzed n-gram use on the basis of statistical association scores. Results from n-gram studies have found significant differences between L1 and L2 writers and between intermediate and advanced L2 writers in terms of their bigram use. The current study expands on this research by investigating the connection between bigram and trigram association measures and human judgments of L2 writing quality. Using multiple statistical association indices, it examines bigram and trigram use by beginner and intermediate L1 Korean learners of English in English placement test essays. Results of a logistic regression indicated that intermediate writers employed a greater number of strongly associated academic bigrams and spoken trigrams. These findings have important implications for understanding lexical development in L2 writers and notions of writing proficiency.