The purpose of this study is to explore the rater reliability and test validity of SEPT (Spoken English Proficiency Test) developed by YBM Sisa-Yong-A-Sa in Korea. For this purpose. the performance of 47 test-takers was rated and analyzed using Rasch modelled FACETS (Linacre & Wright. 1993), Speaking performance in SEPT was also compared with candidates' TOEIC sub-scores. After exploring the psychometric extent to which the test criteria and tasks of the SEPT are valid. it was concluded that the testing scheme partially met the theoretical and empirical consideration behind the test specification and rater training. Implications for further revision of the SEPT are discussed.
Abstract
1. 연구의 필요성 및 목적
2. 연구자료 분석방법
3. 연구 기획
4. 연구결과 및 분석
5. 결론
참고문헌