The structural sensitivity of elicited imitation as a measure of implicit grammatical knowledge

One of the major issues in SLA is developing language tests that could produce good measures of implicit knowledge. This study examines the validity of an English Elicited Imitation (EI) test as a measure of L2 implicit grammatical knowledge. Forty freshman university students in Turkey took a set of language tests: an EI test, two storytelling tasks, a picture description task, IELTS listening sample tests and a speaking test. Four English morphemes were chosen as the target structures: third person ‘-s’, plural ‘-s’, simple past ‘-ed’, and comparative ‘-er’. Results from a principal component analysis showed that all measures were loaded on a single component labelled as implicit knowledge. Significant correlations with varying magnitude were also recorded between learners’ EI scores for the target structures and their scores on other time-pressured measures: r = .63, r = .63, r = .65 and r = .43, for third person ‘-s’, simple past ‘-ed’, plural ‘-s’ and comparative ‘-er’, respectively. These findings suggest that the likelihood of EI measuring L2 implicit grammatical knowledge may vary depending on language structures.