Modelling School Factors and Performance in Mathematics and Science in Kenyan Secondary Schools Using Canonical Correlation Analysis

Abstract

The level of performance and participation in Science, Technology, Engineering and Mathematics (STEM) career subjects remains low in Kenya despite STEM’s critical role in economic development. Numerous factors contribute to students’ academic achievement in STEM education.There is a need to focus on the contribution schools make in assisting students to achieve better scores in STEM. Due to the challenges of poor performance in these subjects, this study endeavors to establish the relationship between the school characteristics and academic achievement in STEM education.The objectives of the study include determining: the magnitude of the relationship between school factors and performance in STEM education, the most influential subject in describing the level of STEM education, the most contributing school factor to STEM education and a model to predict performance in STEM education given school factors. The research utilized data from 9,834 candidates of year 2015 Kenya Certificate of Secondary Education (KCSE) from 77 public secondary schools in Nairobi County. Canonical Correlation Analysis (CCA) is a multivariate data analysis technique that seeks to establish whether two sets of variables, are independent of each other. Given that the two sets of variables are dependent, CCA is able to represent a relationship between the sets of variables rather than individual variables. From the 2015 KCSE data, CCA revealed that school factors significantly correlate with the level of performance in STEM education. Based on standardized canonical coefficients and canonical loadings, the subjects that mainly influence the level of performance in STEM education were found to be mathematics and physics. Further assessment of the canonical cross loadings from the two variate pairs revealed that the proportion of students with mean grades of C+ and above and the proportions of students taking biology and physics contribute very highly to the level of performance in STEM education.