Masood Siyyari - PhD Assistant Professor Science and Research Branch, Islamic Azad University, Tehran, Iran
Setare Ghorban Daei - South Tehran Branch, Islamic Azad University

One variable as a source of error in self/peer-assessment which has not been investigated much in the literature is the purpose of self/peer-assessment, which means whether the students self/peer-assessment accuracy are going to be considered as a part of their final score (i.e. summative purpose) or not (i.e. formative). Given this point, this study investigated whether self-assessment for formative purposes versus summative purposes has any differential effect on the writing performance and the self-rating accuracy of Iranian EFL learners. Sixty homogenized intermediate Iranian EFL learners were assigned randomly to the two assessment conditions. After the administration of a paragraph writing pretest, the ESL composition profile (Jacobs et al., 1981) was introduced to the students for self-assessment purpose. In the formative group self-assessment was practiced by the students during the term after each writing assignment. By subtracting the students self-assessment scores from the raters criterion scores, self-assessment error scores were worked out, and the students could see their accuracy extent in their self-assessment . In the summative group, similar treatment was given except that the students accuracy of self-assessment over the treatment sessions was considered in their final term score. Finally, to check the improvement of the participants in writing and rating accuracy, after the ninth session a posttest self-assessment was administered. The ANCOVA results revealed that under both summative and formative assessment conditions, equally significant improvement occurred in the writing performance of the students; however, in the summative assessment group the students improved much more significantly in self-rating accuracy.

Summative Assessment; Formative Assessment, Writing Ability, Rating Accuracy