The study of the mode effect between computer-based and paper-based science tests in TIMSS 2019

This study analyzed the science assessment items that were conducted by the paper and the computer-based assessment in the TIMSS 2019 pre-test, which is an international academic achievement assessment. We examined whether the test mode influenced the percentage of items answered correctly by each graders. As a result, there was no significant difference in the correct answer rates of science items according to the test media in 4^thand 8^thgraders. Looking at the each graders, the difference in the percentage of correct answers was relatively large in the fourth grader rather than the eighth grader. In terms of items, the difference in the percentage of correct answers was relatively larger for the constructed items than the multiple choice questions. As a result of analyzing the content of science and assessment items on the influence of the test media, the multiple choice items showed little difference between the paper and the computer-based assessment items, but the answertype items tended to have a lower percentage of correct answers than the paper-based assessment. In addition, by grade, 4^thgraders showed lower percentages of correct answers in the constructed items than 8^thgraders. This study provides implications related to the development and introduction of computer-based assessment in Korea, and suggests an educational implications for computerbased assessment to be well established as school evaluation.