This paper explores rating performance of high school EFL teachers in national
curriculum-specific writing assessment by comparing rater reliability and ratings
of experienced and novice raters. The experienced raters completed the
government-authorized writing test rater certification program and had two- or
three-year rating experience, whereas the novice raters were new teachers who
lacked rater training and rating experience. The two groups of teacher-raters rated
410 samples written by high school students in four assessing aspects: task
completion, content, organization, and language use. Results reveal the discernible
influence of rater background on ratings. The experienced teacher-raters’ rating
experience and understanding of the target test task and the rating rubric acquired
through the rater certification process resulted in better rater reliability and
significantly lower ratings than the novice teacher-raters. Variations among
assessment aspects were also noted: for example, lower reliability in content and
organization and lower ratings of the experienced teacher-raters in organization
and language use. The findings from the study imply that test-takers may receive
different ratings depending on who rates their writing, and thus, securing credible
experienced raters is essential in a high-stake writing test.