Using sentence processing speed and automaticity to predict L2 performance in the productive and receptive tasks

Abstract

Abstract We investigated the predictive capacity of second language (L2) sentence processing speed measured using reaction time (RT) and automaticity captured using the coefficient of variation (CV) on the accuracy of L2 use in receptive and productive tasks. A total of 74 Chinese English-L2 learners completed a maze task, an elicited imitation task (EIT), and a timed grammaticality judgment task (TGJT) using comparable sentences. RT and CV from the maze task were used to predict the use of L2 morphological markers in the EIT and TGJT. The findings revealed that there were no significant effects of RT and CV in the EIT. However, in the TGJT, the effect of CV was significant. These results highlight the predictive power of L2 processing automaticity in the receptive task involving L2 morphological use.