이항모형방법과 IRT 방법으로 추정된 조건부 측정오차의 실증적 비교 = Comparison of Conditional Errors of Measurement Using the binomial method and IRT method

다국어 초록 (Multilingual Abstract)

This paper describes two methods previously developed for estimating conditional standard errors of measurement: the binomial method and the IRT method. These methods are based on different underlying assumptions. The binomial method employs the binomial model as the distribution of errors and the IRT method is based on the unidimensional IRT model assumption. Two methods are compared using real data that were gathered in America.
Followings were found: CSEM results from the two methods were similar and error variances could be overestimated in the binomial method when item difficulties in a test were rather different. Based on results, it could be concluded that when IRT parameters were estimated for test development purpose CSEM estimated by IRT method is worth considering. However, when internal consistency reliability such as KR-20 and standard error of measurement were reported, CSEM estimated by the binomial method would be recommended.

참고문헌 (Reference)

1 성태제, "현대교육평가 3판" 학지사 2010

2 성태제, "문항반응이론의 이해와 적용" 교육과학사 2001

3 한국교육평가학회, "교육평가 용어사전" 학지사 2004

4 Kolen, M., "Test equating, scaling, and linking" Springer-Verlag 2004

5 Lord, F. M., "Statistical theories of mental test scores" Addison 1968

6 AERA, "Standards for educational and psychological testing"

7 Lord, F. M., "Standard errors of measurement at different score levels" 21 : 239-243, 1984

8 Feldt, L. S., "Some relationships between the binomial error model and classical test theory" 44 : 883-891, 1984

9 Lee, W, "Personal communication"

10 Kolen, M., "POLYCSEM Manual"

11 Hambleton, R. K, "Item response theory: Principles and Applications" Kluwer․Nijhoff Publishing 1985

12 Embretson, S. E, "Item Response Theory for Psychologists" Lawrence Erlbaum Associates 2000

13 Allen, M., "Introduction to Measurement Theory" Brooks/Cole Publishing Company 1979

14 Lee, W., "Estimators of conditional scale-score standard errors of measurement: A simulation study" 37 (37): 1-20, 2000

15 Kolen, M., "Conditional standard errors of measurement for scale scores using IRT" 33 (33): 129-140, 1996

16 Lord, F. M, "Comparison of IRT true-score and equipercentile observed-score equating" 8 : 453-461, 1984

17 Zimowski, M. F., "BILOG-MG: Multiple-group IRT analysis and test maintenance for binary items" Scientific Software International 1996