<![CDATA[The ability of students to achieve learning objectives can be measured by conducting an assessment. There are various forms of assessment, and the assessment in the form of a testlet is believed to have advantages in identifying the students' conceptual understanding achievements. This study aims to develop a testlet model physics instrument that is valid, reliable, and feasible to use. The development procedure used is 4D which consists of 4 stages, namely: Define, Design, Develop, and Disseminate. However, this research is reduced to 3D only limited to the Develop stage. The research was conducted in a high school located in the city of Semarang. The small-scale trial sample involved 20 students who were selected using a random sampling technique. The results showed that the content validity analysis by the expert obtained an average result of 0.9 in the valid category. The results of the small-scale product test analysis showed that 92% of the items were declared valid and 8% of the items were declared invalid. The question is declared to have very high reliability with a value of 0.974. Analysis of the level of difficulty obtained criteria for 8% difficult questions and 92% moderate questions. The discriminating power of the questions shows the percentage of 2% of the questions being rejected, 2% of the questions being corrected, 2% of the questions being accepted well with improvements, and 94% of the questions being well-accepted.]]>
Copyrights © 2021
