Modeling authentic assessment: Using analysis based on the Structural Equation Model (SEM)

Abstract

This article introduces an authentic assessment model that is built on the influence of student attitudes towards authentic research, while student attitudes can be influenced by factors such as parental influence, teacher ability, and peer influence. Then the attitude of these students is also a factor that affects student learning outcomes using authentic assessment models. This study aims to determine the suitability of the structural equation model with themodel authentic assessment with the variables of student attitudes, peer attitudes, teacher abilities, and understanding and expectations of parents related to authentic assessment and to determine the influence of parents, the ability of teachers, and peers on attitudes. students towards authentic assessment. Want to know the effect of student attitudes on authentic assessment on learning outcomes. And want to know the relationship between parents, the ability of teachers, and peers related to authentic assessment. This research is a quantitative study with a correlational method. The sample of this research was seventh grade students from 14 State Junior High Schools in Cirebon City, West Java Province, Indonesia. The number of research samples was 361 students. The data analysis technique uses SEM (Structure Equation Modeling). The research instrument was a questionnaire of students' attitudes and perceptions in authentic assessment and mathematics learning outcomes. The questionnaire instrument of students' attitudes towards authentic assessment was 54 items and the mathematics learning outcomes test consisted of 18 multiple choice items and 6 items. The research instrument before being tested was tested for content validity and construct validity. Analysis of the test instrument using the Rasch model, the results are that the two research instruments are stated to have a fit model and good reliability levels. The results stated that there was a match between the range model and the research result model, there were 2 hypotheses rejected, namely hypothesis 2 and hypothesis 5, namely there was no influence of peer attitudes and student achievement related to authentic assessment and there was no influence of peer attitudes on student achievement results. related to authentic assessment. Regarding the rejection, because the data was taken when learning was carried out online during the pandemic, therefore the peer variable had no effect on student attitudes or on student learning achievement related to authentic assessment.