The relationship of lexical proficiency to the quality of ESL compositions

Abstract

The extent to which impartial readers take into account lexical richness and lexical errors when assigning a quality score to compositions written by learners in an intensive English program is discussed in this article. For placement purposes into both ESL programs and academic programs, the writing of these students is often assessed by anonymous readers who base their judgments on timed writing tasks. Much remains to be known, however, about the relationship between language proficiency, specifically lexical proficiency, and reader judgments of the overall quality of timed essays. This study reports on the role of the lexical component as one factor in holistic scoring. Sixty-six placement essays written by students from mixed language backgrounds in the intermediate to advanced range of an intensive English program were holistically scored. These quality scores were then compared to four lexical richness measures: lexical variation, error-free variation, percentage of lexical error, and lexical density. High, significant correlations were found for (a) lexical variation, that is, the ratio of the number of different lexical items to the total number of lexical items in the essay adjusted to length; and (b) lexical variation minus error. The latter measure, error-free variation, correlated best with score.