Using Multilevel Modelling to Evaluate Science Literacy and Technology Course of the Indonesian Non-Science Students

Abstract

In this paper, a science literacy and technology course has been designed and implemented to strengthen the national initiative empowering scientifically literate Indonesian society. This paper is intended to evaluate to what degree non-science undergraduate students can perform this course. The diverse background of non-science students who participated in this study led to the challenge to evaluate their performance more comprehensively contemplating the nested structure of students’ department and faculty setting. In light of the hierarchical nature of the student data, multilevel modelling was used to conduct the analysis. The first level of analysis involved students’ performance and affective attributes measured using demonstrated science literacy assessment (SLA-D) and motivational beliefs (SLA-MB) respectively. Then, the subsequent level of analysis comprised demographic factors gathered from the institutional record. Findings demonstrated that the impact of demographic factors on the students’ performance of science literacy was not substantial. Different settings of students’ department and faculty level drove the association between affective factors and the learning process toward science literacy courses substantially. The multilevel approach controlled the equitable student assessment within the nature of students’ data structure. This paper suggests an implication of advancement regarding educational data analysis and examines the effectiveness of science literacy courses for higher institutions specifically for non-science majors.