Abstract
Education plays a fundamental part in self-development since it capacitates individuals with cognitive, procedural, and attitudinal competences, thus promoting employment, earnings, and health. For societies, it drives economic growth, spurs innovation, and fosters social cohesion and poverty reduction. Research has increasingly dwelled on understanding the drivers of academic achievement to better inform decision-makers. Educational policies in recent decades in Portugal have increased literacy rates, yet the number of dropouts before high school (compulsory education) completion is still a matter of concern when compared to the European graduation average rates. The present thesis aims at shedding light on what are the main determinants of academic achievement in Portuguese public secondary education. By combining various data sources – literature review, primary and secondary data – and employing robust data analysis methods – structural equation modeling and machine learning techniques – this work identifies and classifies, by order of importance (i.e., which of them has a greater impact on students’ grades), determinants pertaining to students, teachers, and parents. Results allowed for a comprehensive overview of the determinants of academic achievement, highlighting the more significant ones. The two main drivers of success are previous retention and parental education level. Specifically, having failed a year in school impairs high school achievement, while having a legal guardian who holds a university degree promotes it. This result was found in both primary and secondary research, thus eliciting that combining multiple data sources and resorting to sophisticated analysis methods poses an advantage in this field of research. Moreover, it has significant practical implications, as it helps to advise policies in order to increase graduation rates.
Original language | English |
---|---|
Qualification | Doctor of Philosophy |
Supervisors/Advisors |
|
Award date | 18 Jul 2023 |
Publication status | Published - 18 Jul 2023 |
Keywords
- Academic achievement
- High school
- Structural equation modeling
- Machine learning
- Sucesso Escolar
- Ensino Secundário