The item response theory(IRT) is an analytical method that makes it possible to objectively inspect the characteristics of items and that makes it possible to exam the non-objectivity possessed by a number of sources, in which they would check validation by language assessment research. In this paper, item response theory analysis were conducted on the large scale implementation result for the Korean Speaking Test for Academic Purposes(KoSTAP). The test results were constructed systematically so that the meaning of the IRT analysis can be seen in the test and scoring process. For the analysis of the results, General Partial Credit Model(GPCM) is selected and the effect of the rater is considered in the characteristics of the performance test. Based on the analysis, the characteristics of KoSTAP were investigated mainly on parameters for item, rater, and person. Finally, with the test results, we looked at whether KoSTAP was objective characteristics and presented an ideal direction to improve its validity as a large-scale high-stakes test.
1. 서론
2. 선행 연구
3. 연구 방법
4. 연구 결과
5. 논의
6. 결론
참고문헌