Correction Effects of Response Bias of Different Types of Score Adjustment in the Anchoring Vignette Method

서지정보

· 발행기관 : 한국교육학회
· 수록지 정보 : 교육학연구 / 59권 / 7호 / 25 ~ 53페이지
· 저자명 : 조시정, 백순근

초록

The anchoring vignette method is a popular method for correcting response bias in Likert-type measurements. It is a technique of correcting an individual's response to a measurement item by using the responses to vignette items on the same variable. However, out of ways one can adjust respondents’ scores using anchoring vignettes, comparisons across different methods of score adjustment have been insufficiently discussed. Hence, this study analyzed and compared the correction effects of response bias of different methods of score adjustment using anchoring vignettes in the measurement of achievement motivation from the Programme for International Student Assessment (PISA) 2015 data. Four different methods of score adjustment － (a) using the upper bound of the interval score, (b) using the lower bound of the interval score, (c) using the median of the interval score, and (d) using the value suggested by the Organisation of Economic Cooperation and Development (OECD) － were analyzed and evaluated through three assessment criteria (direction of correction effect, size of correction effect, and robustness against rating order) for their suitability. Findings indicate that there are differences in the correction effects of response bias across these different methods of score adjustment and that using the median of the interval score yielded the most logical and robust correction effects of response bias. These results suggest that we should be careful in selecting a method of score adjustment when using the anchoring vignette method.

참고자료

· 없음