Measuring the contribution of academic and general vocabulary knowledge to learners' academic achievement

Publication Type:

Journal Article


Journal of English for Academic Purposes, Volume 31, p.44 - 57 (2018)




Academic vocabulary, Rasch model, Receptive knowledge, Test validity, Vocabulary size


Abstract The Academic Word List (AWL) (Coxhead, 2000) is widely used in preparing non-native speakers for academic courses, and it is thought that the words in this list are essential for the understanding of English academic texts (Cobb & Horst, 2004). It is also thought that the \{AWL\} is a list of infrequent and specialised words inaccessible from general language. These preconceptions are challenged in the current study. With reference to BNC/COCA word lists, the study demonstrates that the majority of the words from the \{AWL\} fall within the 3000 most frequent words, a grouping that Schmitt and Schmitt (2014) describe as highly frequent. Using a specifically created test of the \{AWL\} and a test of overall vocabulary size (XK-Lex; Masrai & Milton, 2012), the study demonstrates that the learning of the \{AWL\} appears to be strongly influenced by the frequency of these words in general corpora and that the \{AWL\} test very strongly resembles a test of overall vocabulary size. Knowledge of the \{AWL\} also adds marginally to the power of overall vocabulary size in explaining variance in grade point average (GPA) scores. This conclusion matches that of Townsend, Filippini, Collins, and Biancarosa (2012), although the tests in the current study appear to have greater explanatory power.