Dropout in rural higher education in virtual modality: a neural network approach

Access to higher education is a cornerstone for developing nations and individuals, as it contributes to economic growth, the reduction of inequality and improvements in quality of life. To address educational disparities in rural areas, governments have implemented various strategies to ensure the right to higher education, with virtual education emerging as a significant approach. However, despite these efforts, student dropout rates in virtual programmes remain a considerable challenge, especially in rural contexts where socioeconomic and academic conditions are often unfavourable. This study aimed to identify and predictive patterns associated with dropout among higher education students variables enrolled in virtual undergraduate programmes in rural areas. The study was developed with data from students belonging to educational programmes in Colombia, with a sample of 269 rural students. Using artificial neural networks (ANNs), the study analysed the influence of variables grouped into individual, socioeconomic, academic and institutional determinants on the dropout intentions of rural students in virtual programmes. The findings provided a comprehensive view of this complex phenomenon, highlighting the interaction between multiple determinants and offering tools to design strategies that mitigate dropout rates and enhance retention among rural students.