Sentence repetition (SR) tasks are widely used to assess syntactic knowledge in populations with and without language disorders. The present study investigated if the use of several scoring methods would provide insights into the validity of a researcher-created SR task as a measure of syntactic knowledge.
103 grade 6 students (˜12 year olds) from ordinary Danish classrooms completed the task, which consisted of 30 reversible sentences. Items with canonical subject-verb-object (SVO) word order were included along with sentences with non-canonical word order, e.g. passives and object-verb-subject (OVS). Responses were scored for 1) accuracy, 2) preservation of syntactic structure and 3) change of syntactic construction to SVO.
Canonical constructions were repeated more accurately than the non-canonical constructions. However, whereas the syntax of the passives was largely preserved even in inaccurate sentences, OVS items were vulnerable to syntactic change to canonical SVO.
These differences reflect the dimensionality of the syntactic construct and thus indicate that the present SR task measures aspects of syntactic knowledge.
Funding Sources: Independent Research Fund Denmark and University of Copenhagen.