NotesFAQContact Us
Search Tips
ERIC Number: ED504423
Record Type: Non-Journal
Publication Date: 2009-Apr-1
Pages: 26
Abstractor: As Provided
Reference Count: 22
ISSN: ISSN-0974-8741
Identifying Sources of Bias in EFL Writing Assessment through Multiple Trait Scoring
Salmani-Nodoushan, Mohammad Ali
Online Submission, The Modern Journal of Applied Linguistics v1 p28-53 Mar 2009
For purposes of the present study, it was hypothesized that field (in)dependence would introduce systematic variance into EFL learners' performance on composition tests. 1743 freshman, sophomore, junior, and senior students all majoring in English at different Iranian universities and colleges took the Group Embedded Figures Test (GEFT). The resulting 582 Field-Independent (FI) and the 707 Field-Dependent (FD) students then took the 2000 version of the IELTS. Using SPSS commands for collapsing continuous variables into groups, and participants' IELTS scores (based on 25, 50, 75 percentiles), four proficiency groups were identified for each kind of cognitive styles. From each proficiency group, 36 FD and 36 FI individuals were selected through a matching process. The scores obtained by the resulting sample of 288 participants on the second writing task of the IELTS test were used as the data for this study. The results of data analysis revealed that individuals' cognitive styles resulted in a significant difference in their writing performance in proficient, semi-proficient, and fairly proficient groups, but not in the low-proficient group. The findings also indicated that cognitive style resulted in a significant difference in participants' performance on such aspects of EFL composition as content, structure, and language. (Contains 2 figures and 8 tables.)
Publication Type: Journal Articles; Reports - Research
Education Level: Higher Education; Postsecondary Education
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Identifiers - Location: Iran
Identifiers - Assessments and Surveys: International English Language Testing System