This study aims to explore the scientific creativity of high school students on the static fluid in STREM PBL with an e-authentic assessment. This study used a mixed method with an embedded experimental design. This study involved 30 11th-grade senior high school students in Sidoarjo, Indonesia. Data collection techniques were carried out through pretest-posttest, interview, and observation. Pretest and posttest instruments consist of three scientific creativity essay questions with a reliability of 0.610. The results of the Wilcoxon test of p = .000, with the posttest being higher than the pretest, indicate a significant difference between students’ scientific creativity before and after learning. The effect size value of 0.87 indicates that STREM PBL with an e-authentic assessment moderately affects students’ scientific creativity. All indicators of scientific creativity have increased. The order of increasing the average scientific creativity score for each indicator is fluency (high) elaboration (moderate) originality (moderate) flexibility (moderate). Meanwhile, the order of the average level of scientific creativity at the posttest is fluency (very creative) elaboration (quite creative) = flexibility (quite creative) originality (less creative). Students’ scientific creativity increases because this learning makes students solve problems, carry out projects (design and create a product), carry out inquiry activities (experiment and investigation), collaborate with groups, and evaluate projects through self and peer assessment. Learning activities that can increase the level of fluency indicators are mentioning as many ideas as possible for experimental designs and product designs. Learning activities that can increase the level of flexibility indicators are integrating religion content into STEM (it is also better if integrating art content) and providing several problems and products from various fields. Learning activities that increase originality indicators give each individual or group a different project topic. Learning activities that can increase elaboration indicators are strengthening mastery of concepts and evaluating the work of oneself and others through self-assessment and peer assessment.

Research and development of computer-based assessment is carried out to overcome the shortcomings of conventional paper-based assessment. Development is carried out by focusing on the use of isomorphic questions and improvement programs. This computer-based assessment was developed with two levels of questions. At each level there are isomorphic questions that are used as a reference in providing feedback to students about their concepts understanding. At each level there is also video material that can be used as a source of student learning before students utilize the remedial program. This computer-based assessment has been tested on 31 third-year undergraduate students in Physics Education, Universitas Negeri Malang. The results of media and material validation and the results of limited trials indicate that the computer-based assessment produced is feasible to be applied in learning.

The aim of this study is to analyze the ability of students to solve the problems of linear motion kinematics expressed in symbolic and numeric representation. Research was survey with cross-sectional design. Research subjects included 26 first year undergraduate students in physics at one of the State Universities in Malang which was consisted of 10 men and 16 women. The research instrument was open-ended test of linear motion kinematics problems expressed in symbolic and numeric representations with a reability of 0,807 The research data were analyzed using descriptive and non-parametric inferential statistics. The results showed that the ability of students to solve linear motion kinematics problems in both symbolic and numeric representation was medium. Students had difficulty solving physical problems in both symbolic and numeric representations. It was also found that the problems of linear motion kinematics in symbolic representations were more difficult for students to solve than numeric representations. The study suggested further research to explore the causes of student difficulties more authentically, e.g. by interviewing or thinking aloud.