Development of a Test Instrument to Test Student Comprehension Based on Multiple Representation of Buffer Solution Material Using the Rasch Model

Abstract

The results of students' daily tests should be an evaluation material for teachers to evaluate students' understanding of the three levels of multiple chemical representations. The use of test instruments can be used as future learning so that teachers can better know which level of representation needs to be emphasized in the learning process and teachers can find out students' understanding abilities of the interconnection of these three levels, especially in buffer solution material. The Rasch model is used in this research, which is of the Research and Development (R&D) variety, to create a test instrument that can assess a student's comprehension of three levels of chemical representation in buffer solution content. The test instrument is valid, reliable, has a difficulty index, and has strong item discrimination. The steps of this research raised 10 stages of development by Wei, et al in 2012. The research instruments used were a questionnaire validating the contents of the items and also a questionnaire validating the assessment rubric which was filled out by 5 chemistry lecturers who are experts in their fields. Raw data from lecturer validation results were analyzed using the Rasch model supported by Minifacet and Ministep Software. The exact agreement results were obtained from the results of the content validity of 90.6% and the expect agreement of 91.1%, which means that there is a fit between the results of the lecturer's assessment and the results of the model design. Furthermore, the results of empirical item quality found that all items met the valid criteria on the MNSQ, ZSTD and Pt-MeanCorr indicators with a reliability value of 0.91 in the very good category, had a difficulty index, namely easy, medium and difficult levels, as well as the differentiating power of the questions. has 3 levels of respondents.