| مستخلص: |
The purpose of the current dissertation is to map the relationships between first language (L1), writing quality, and syntactic complexity, accuracy, lexical complexity, and fluency (CALF) in second language (L2) writing. CALF are characteristics of language production that have been of significant interest in L2 writing research for the past few decades. Though they have been extensively studied as dependent variables that may vary as a function of other factors, they have been rarely studied together, much less in relation to L1 as an independent variable. Thus, this study explored the effects of L1 and writing quality, operationalized as score levels, on all four dimensions of CALF and the predictive power of CALF measures on writing quality. Adopting a quantitative, corpus-based approach, I collected 1,683 essays from the Educational Testing Service (ETS) Corpus of Non-Native Written English (TOEFL11) for analysis. The corpus is comprised of essays written by speakers of 11 non-English native languages (Arabic, Chinese, French, German, Hindi, Italian, Japanese, Korean, Spanish, Telugu, and Turkish) as part of an international test of academic English proficiency -- TOEFL (Test of English as a Foreign Language). The selected essays were controlled for topic and collapsed into three score levels: low, medium, and high. They were automatically processed for 14 syntactic complexity measures, five lexical complexity measures, and one fluency measure using different automated tools. Approximately 20% (329 essays) were hand-coded for six accuracy measures. Statistical tests revealed that there were significant differences between L1s in most CALF measures in all score levels. Text length (W/Tx) was found to differentiate score levels in all L1s. Other relatively consistent indicators of score levels across L1s are the total number of errors and lexical diversity measures such as the index of lexical diversity (D) and the measure of textual lexical diversity (MTLD). Multinominal regression models output mean length of sentence (MLS), the number of coordinate phrases per clause (CP/C), lexical density (LD), MTLD, lexical sophistication (LS1), and W/Tx as predictors of high-quality writing. Overall, results showed that CALF measures varied significantly across L1 backgrounds and score levels with several measures being predictive of the writing quality of a heterogenous group of L2 writers. These findings suggest that CALF should be examined together when assessing L2 writing and that L1 background is an important factor to consider when studying CALF in L2 writing. It is also necessary to tailor L2 instruction and assessment to address the unique challenges learners from different L1s face. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.] |
