Year: 2017
Type: dissertation
Abstract: This study investigates the multicomponential nature of L2 Arabic writing by adapting a number of direct measures of linguistic complexity, accuracy, and fluency (CAF) to the measurement of written Arabic texts produced by L2 Arabic learners and Arabic Native Speakers (NSs).Previous studies of L2 writing have established that such measures are reliable indicators of learners' developing interlanguage, however few efforts have thus far applied CAF analyses to spoken or written Arabic.The results of the measures explored here are triangulated with learner responses to several independent estimates of Arabic proficiency, including a novel "short-cut" estimate in the form of an Arabic C-test.Instead of a longitudinal design, in which a learner or group of learners are asked to provide data at different stages of L2 acquisition, the current study entailed the collection Arabic production along a spectrum of writing ability, spanning from learners at the end of their first year through Arabic NSs.Significant findings of this study include the fact that learners appear to "move" toward more natively-like written production as their overall facility with the Arabic language increases.Furthermore, learners scores on the various selected CAF measures correlate highly with each other, as well as with the selected independent measures.C-test consistently emerges via regression as the most reliable predictor of variability in written CAF among the selected explanatory variables, followed closely by self-estimated overall Arabic ability.Conversely, however, this study establishes that sentence-based measures are shown to be unreliable for the
Author: Michael Raish
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W2732710139