نسیم قنبری

As a matter of fact, writing scores have important consequences for test takers where the result impacts the professional and educational aspects of their lives, so the sources of errors especially rater errors is needed to be identified and reduced . For this aim, as one of the sources of rater variability this study investigated the rating behaviors of raters in two different proficiency states. It also looks into the items which raters pay special attention to when they are engaged in assessing EFL writing. For this purpose, ten experienced raters provided their think-aloud protocols (TAPs) while scoring 4 essays, 2 written by BA students and the other 2 belonged to MA students by using ESL Composition Profile (Jacobs et al. 1981). The result of the t-tests showed that there were significant differences between the mean scores of the raters attributed to the BAs’ papers in aware and unaware states; however, there was no significant difference between the mean scores of the raters attributed to the MAs’ papers in aware and unaware states. Students’ proficiency level differences were also found to be effective in the raters’ different rating patterns when rating learners’ writing tasks.