평정자간의 엄격성 차이정도가 피험자 총점산출 방법에 미치는 영향: 원점수, 표준점수, Facet점수 비교

초록

This study investigated the difference between person's raw score and score adjustments using standardized score and facet score under the different conditions of rater's severity. To demonstrate how various rater severity manifest themselves in this analyses, data were simulated to represent five types of rater severity: (1) no effect (2) random (3) leniency (4) harshness (5) wide spread of rater severity. Data generation was performed using WinGen and Facet computer program. The results showed that the agreement between person's raw score and standardized score was high regardless of rater severity conditions. On the other hand, as the magnitude of the rater severity increases, the agreement of raw scores and facet score was dramatically low. The reasons for the results were discussed and suggestions were made for future studies.

키워드