Measures of Syntactic Complexity in L2 Arabic OPIs Across ACTFL's Advanced Level
Author
McCombie, SethIssue Date
2022Keywords
ACTFLArabic L2 acquisition
Linguistic complexity
OPI
Proficiency thresholds
Syntactic complexity
Advisor
Azaz, Mahmoud
Metadata
Show full item recordPublisher
The University of Arizona.Rights
Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction, presentation (such as public display or performance) of protected items is prohibited except with permission of the author.Abstract
An L2 learner’s language proficiency is composed of several facets such as linguistic complexity, fluency, and accuracy (CAF). Measuring these facets across proficiency levels can be challenging. Perhaps as a result, even widely adopted oral proficiency scales lack clear, operationalized definitions of these three constructs. To highlight the need for more empirical and granular profiles of oral proficiency in terms of these constructs, I discuss some possible deficiencies in the oral proficiency guidelines of the American Council on the Teaching of Foreign Languages (ACTFL). I then attempt to create a clearer profile of Advanced-level speech in terms of one aspect of proficiency-- syntactic complexity (SC). To do this, I analyze 16 officially rated OPIs with ratings ranging from Intermediate-high (IH) to Advanced-high (AH) in terms of several metrics of syntactic complexity to identify if any measures of SC were predictive of a student’s proficiency level. This analysis confirmed that the ratio of subordinate clauses to independent clauses in an interviewee’s speech is significantly correlated with his/her proficiency ratings at or near the Advanced level (p = .05). Other measures of syntactic complexity, such as the use of subordinate clauses in isolation and the length of subordinate clauses were not predictive of a student’s proficiency rating. The latter findings confirm the multiplex nature of complexity and underscore the need to consider multiple facets of proficiency together. It further indicates that distinctions between sublevels of the ACTFL scale may not always be statistically significant or even apparent when measured in terms of one metric of complexity in isolation. Ultimately, I suggest additional research directions and pedagogical implications that could help students further develop their SC and help researchers assess its connection to proficiency ratings. I do this with an eye toward the future of automated proficiency scoring in Arabic in the coming years.Type
textElectronic Thesis
Degree Name
M.A.Degree Level
mastersDegree Program
Graduate CollegeMiddle Eastern & North African Studies
