Developing an assessment instrument based on HOTS according to Brookhart for thermochemistry materials

Abstract

The Indonesian Ministry of Education and Culture has developed a learning assessment system in the 2013 curriculum that applied higher-order thinking skills in order to foster those skills on Indonesian students albeit the paramount importance of these skills, application assessment is still low in reality. It can be seen from the results of research on HOTS assessment applications by teachers in several high schools in East Java in the formatives and summative examination questions, which still range from 9-13.39%, which is relatively small. Therefore, the development of HOTS instrument is integral as a model of HOTS assessment for chemistry teachers and students. In the 2013 Curriculum, reference of the HOTS development is delivered by Brookhart, which consist of transfers, judgment, logic and reasoning, creative thinking, and problem solving. Thermochemistry topic was chosen because the topic has diverse conceptual, algorithmic, and contextual aspects, and they are interconnected among these aspects, so that the HOTS instrument is qualified to apply. Furthermore, the assessment instrument is called a Brookhart’s HOTS Instrument of Thermochemistry. The research and development of it uses the Design and Development Research (DDR) method according to Richey and Klein in four stages. Instrument validation is comprised of logical and empirical validation. Logical validation includes content and construct validation and was conducted by two expert lecturers whereas the empirical validation was conducted on 87 students of grade XI. Thirty questions were constructed for the instrument. Based on the results of content and construct validation, they are very valid with a percentage of 89% and validity range from 0.114-0.621 (26 items are valid with high reliability score of 0.702 and 4 items are invalid). The difficulty level analysis shows that there are 2 easy questions, 17 moderate questions, and 7 difficult questions, and the discriminating power analysis shows that there are 12 very good questions, 8 reasonably good questions, 2 marginal questions, and 4 poor questions.