Automatization in second language sentence processing: A partial, conceptual replication of Hulstijn, Van Gelderen, and Schoonen's 2009 study

Abstract

ABSTRACTThis study investigates the automatization of sentence processing using the coefficient of variation (CV), a measure of intraindividual processing stability (Segalowitz & Segalowitz, 1993). A smaller CV (i.e., standard deviation to reaction time [RT] ratio) and a positive CV-RT correlation are taken to index increased automatization. Hulstijn, Van Gelderen, and Schoonen (2009) were the first to try to validate the use of CV at the sentence level; however, they did not find any evidence for automatization in their CV analyses. Forty Korean English as a second language students (20 intermediate, 20 advanced) and 20 native speakers performed three speeded tasks in English: a semantic classification task, a sentence verification task, and a sentence construction task. The results revealed that, consistent with findings from previous word recognition studies, the CV in the sentence-level tasks decreased as participants’ proficiency level increased. Although the CV-RT correlation in the sentence verification task was not always significant, no counterevidence against Segalowitz and Segalowitz’ (1993) hypothesis was found. The sentence construction task discriminated better between groups than the sentence verification task. We argue that the CV may be a valid measure of automatization at the sentence level, provided the tasks used target lower-level processes such as word recognition, parsing, and semantic proposition formation.