Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.486367
Title: A longitudinal study of vocabulary in L2 academic English writing of Arabic first-language students: development and measurement
Author: Turlik, Jan
ISNI:       0000 0001 3537 8896
Awarding Body: University of the West of England
Current Institution: University of the West of England, Bristol
Date of Award: 2008
Availability of Full Text:
Access from EThOS:
Abstract:
This corpus based study investigates the lexical development of forty-two 'English as a Second language' (ESL) university students, using 340 formal examination essays, written at regular intervals over a period of twenty-seven months. One of the main research questions is how vocabulary growth during this twenty-seven month period can be modelled. It was assumed, after Laufer (1994), thjilt the growth of basic words would . follow a curvilinear pattern and flatten out towards the end of the period, whereas the learning curve for advanced words would follow a linear pattern with constant growth, continuing perhaps even at higher proficiency levels. A further question was whether, and to what extent, trained and experienced English as a Foreign Language (EFL) teachers (and International English Language Testing System (IElTS) raters) would base their ratings of the essays on vocabulary richness. Earlier research (Malvern & Richards 2002) indicates that teacher judgements in oral interviews might not be influenced by the lexical diversity of the candidate's performance. The question is whether this might also apply.to written texts. The handwritten essays were all transcribed, verbatim, and prepared for processing within predetermined parameters. Measures of development included the use of advanced words (AWL), Malvern and Richards' lexical diversity D (LDV) (requiring the texts to be put into the CHAT format), P_Lex (PL), Guiraud (G), Guiraud Advanced (GA) and Limiting Relative Diversity (LRD). Several regression analyses show that almost all measures display a similar pattern of development and log-linear fitted lines did increase the explained variance slightly. Against the general pattern of vocabulary development, the study also examines individual differences and finds that these do not conform to the same pattern as the group, which leads to a high amount of unexplained variance in the regression analyses. Lexical development as shown by the measures was compared to raters' assessments of vocabulary. Holistic ratings were also given in an attempt to gauge deployment and use of the vocabulary, and some significant correlations were found between the ratings and some measures. This suggests that raters are influenced by other factors such as organisation and grammar and that rating is not based on lexical richness, which agrees with the findings on oral interviews, found in the literature. A high correlation between the lexical and holistic ratings suggests perhaps that raters do not judge vocabulary richness separately but give only holistic judgments.
Supervisor: Not available Sponsor: Not available
Qualification Name: University of West of England, 2008 Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.486367  DOI: Not available
Share: