Developing Dropout Predictive System For Secondary Schools, By Using Clasification Algorithm: A Case Study Of Tabora Region

ABSTRACT 

Recently, there has been an increase of enrollment rate of secondary schools students in Tanzania, due to introducing fee free and expansion of compulsory basic education from pre-primary to form four. These efforts aimed to alleviate poverty and develop national economy by enhancing competitive labor skills. However, the completion rate of students is highly affected by extreme dropout rate in those secondary schools. Previous studies have explored the causes of school dropout problem and they came with recommendation based on treatment measures while this study deals with preventive measures to the dropout problem. Therefore, this study targets to develop dropout predictive system for secondary schools by using classification algorithm. The study guided by system theory on identifying the best predictive features, which cause student dropout. A sampled students drawn in four councils of Tabora region by using purposive and non-probability sampling techniques during April to May 2019, an exploratory sequential mixed method, questionnaire and documentary review methods used to collect data. Tabora region was considered as the region with the highest school dropout rate in Tanzania. The results indicate that distance to school, time used by a student to school, guardian living with student and student resident as a social factors, and performance scores in standard four and six as academic factors that have a strong impact to the targeted variable dropout time. The developed system predicts significant time and class of student being to drop. The study recommends the use of dropout predictive system, that could identify students who at risk of dropout at first day of their registration.