Dissertation Defense: Lexico-grammatical Complexity in EAP Student Writing: A Learner Corpus Analysis
December 13, 2017 @ 01:00 PM - 03:00 PM
Announcing the Final Examination of Ekaterina V. Goussakova for the degree of Doctor of Philosophy
The goal of this descriptive linguistic study was to explore written language development as a result of student enrollment in the English for Academic Purposes (EAP) program at a large state college. Lexico-grammatical complexity (as measured by nine phrasal and clausal features plus type to token ratio [TTR], average word length, and word count) of placement and exit student writing samples was analyzed using corpus linguistic methods. Writing samples (n = 258) were typed, matched, and tagged. A concordance software was employed to produce lexical realizations of grammatical features.
Previous research has shown that academic writing was not as grammatically "complex" as previously thought; however, certain embedded phrases and nominalization as well as nouns in all positions were indicative of grammatical complexity in L2 writing. On the other hand, adverbial clauses, previously considered as indicators of increased complexity and as a result L2 proficiency, were indicative of earlier stages of language development. (Biber et al., 1998, 2011, 2016; Bulté & Housen, 2014; Granger & Paquot, 2008; Staples & Reppen, 2016; Taguchi et al., 2013).
In this study, paired samples analyses showed that frequency counts for such features as pre-modifying nouns, attributive adjectives, adverbial conjunctions, coordinating conjunctions, TTR, average word length, and word count changed significantly, and students produced more of those features in their exit writings than in their placement essays. The frequencies of verb + that clauses and subordinating conjunction because, though non-significant, actually decreased.
A split plot ANOVA analyzed whether a change could be attributed to grammar instruction in EAP 1560, and the results were mixed. Five multiple linear regression analyses were conducted to predict frequencies of exit pre-modifying nouns, attributive adjectives, noun + that clauses, adverbial conjunctions, and TTR from EAP 1560 and EAP 1640 grades, LOEP scores, and the number of semesters students spent in the EAP program. The only significant regression analysis was with TTR, and 28% of its variance could be explained by the independent variables.
Overall, the results indicated that compressed phrasal features were indicative of higher complexity and EL proficiency, while clausal features could have been acquired earlier and signaled elaboration, as previously described in peer-reviewed literature.
Committee in charge: Keith Folse (Chair), Randi Reppen, David Boote, Eleanor Witta