Empirical Testing of a Multidimensional Model of School Dropout Risk

Abstract

Education systems are working to reduce dropout risk, thereby reducing early leaving from education and training rates (ELET) for a more sustainable society. There is a wealth of research on the causes of dropout risk, but little that looks at it in a complex way. Previous research has typically examined the association of a single factor with school dropout. This paper aims to examine the collective relationship between individual, family, and school-level factors and dropout risk based on international literature. Our analyses are based on two online surveys that were conducted among teachers and students in the 2018/2019 and the 2019/2020 academic years respectively (using the data of 2649 students and 2673 teachers from 149 schools in total). Multiple linear regression analyses were performed, and the (ordinary least squares—OLS) regression models were built hierarchically (blockwise entry) with the ENTER method. The research question was which factors are more likely to predict dropout risk. The findings reveal that individual and family factors are far more strongly associated with students’ dropout risk than school-related factors. The two strongest individual factors are learning engagement and performance-oriented learning School factors hardly have a role in preventing dropping out of school. Four of the school factors appear to have a definite effect: standards set for students and teachers, belief in the school’s role to compensate for disadvantages, and positive school climate. All this draws the attention of practising teachers, school leaders and educational policymakers that the school’s protective factors should be stepped up, and the preventive intervention should focus primarily on these factors.

Keywords

• primary education
• early school leaving
• ELET
• school dropout risk
• online survey
• complex model