Assessment Behavior and Perceptions of Raters in Paired and Group Oral Interaction
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The study considers the assessment of L2 English learners by trained raters in paired and group oral assessments in comparison to an individual, monologue assessment, to determine I) the degree to which raters assign pairs/groups shared (the same) scores and the degree to which raters give individual members of pairs/groups higher or lower as opposed to matching individual scores, 2) assessment tendencies by participants`` English proficiency and paring/grouping conditions, 3) individual variation among raters, and 4) the types of comments (positive, negative, or mixed) raters give participants in pair/group conditions. It was found that 1) on the whole, pair/group scores tend to be shared and lower than individual scores, 2) raters tended to assign both higher- and lower-level speakers lower scores when they are in interaction with lower-level speakers, but to assign middle-level speakers better scores in such a condition, 3) raters exhibited significant inter-rater variability, and 4) raters tended to give more negative comments when assigning lower scores and more positive comments on higher scores.
Abstract
The study considers the assessment of L2 English learners by trained raters in paired and group oral assessments in comparison to an individual, monologue assessment, to determine I) the degree to which raters assign pairs/groups shared (the same) scores and the degree to which raters give individual members of pairs/groups higher or lower as opposed to matching individual scores, 2) assessment tendencies by participants`` English proficiency and paring/grouping conditions, 3) individual variation among raters, and 4) the types of comments (positive, negative, or mixed) raters give participants in pair/group conditions. It was found that 1) on the whole, pair/group scores tend to be shared and lower than individual scores, 2) raters tended to assign both higher- and lower-level speakers lower scores when they are in interaction with lower-level speakers, but to assign middle-level speakers better scores in such a condition, 3) raters exhibited significant inter-rater variability, and 4) raters tended to give more negative comments when assigning lower scores and more positive comments on higher scores.