Structural Equation Modeling of Students’ Performance in Pre-Calculus: Basis for Intervention in Senior High School

Early intervention in education is vital to the progress and improvement of diverse learners that can significantly improve long-term educational outcomes and success in education like identifying and providing immediate intervention to students with learning difficulties, especially in mathematics subjects. Thus, this study aimed to develop Structural Equation Modeling, specifically the direct and indirect relationship among the students’ demographic profile, cognitive support, learning environments, student engagements, and performance in the first quarter in Pre-Calculus among grade 11 STEM students of Notre Dame of Tacurong College and Tacurong National High School, to provide a foundation for intervention in senior high students. The research instrument used in this study was a validated survey questionnaire created by the researcher and grouped into five latent variables, which were composed of thirty-three observed variables. The data was examined using frequency, percentage, mean, standard deviation, causal path analysis, and structural equation modeling. The results revealed that the students are diverse in their demographic profile. The predictors of students’ performance in pre-calculus (cognitive support, learning environment, and student engagement) need a very low to moderate intervention. Meanwhile, the level of the student performance was generally satisfactory. The demographic profile and student engagement have a direct connection with students' performance, whereas the cognitive support and learning environment have no association with students' performance. Thus, the link between cognitive support and learning environment has a direct impact on student engagement outcomes. Furthermore, the results revealed that student engagement mediates the relationship between cognitive support and student performance, demonstrating a substantial relationship. In contrast, there is no mediation effect between students' learning environment and their performance through student engagement. Furthermore, the structural equation modeling of five latent variables resulted in twenty-two indicators that indicate the reasonable level of convergent validity, discriminant validity, and significant relationship. Furthermore, the student engagement and student performance of the model indicate 35.6% and 9.3%, showing statistically significant support as a basis for intervention.