This study utilized a data mining technique (classification tree) to examine the influence of school designation, status, type, size, and student gender on academic achievement. The results from this study show that it is possible to classify whether a student will attain the minimum grade to be qualified for a public university in Kenya with relatively high accuracy. The classification tree correctly classified 93.7% of the students who attained a grade lower than C + and 41.6% of the students who achieved grade C + or a higher grade from the KCSE, resulting in an overall prediction accuracy of 76.7%. The high prediction accuracy for ‘failing students’ is welcome because it gives the stakeholders where work is needed, and where measures to improve the quality of education and enhance resource management should be channeled.
In regard to the most important variable in the classification of whether the student is qualified for a Kenyan public university, school residential status was the most important predictor, but school type, size, and designation were also very important. On the other hand, the gender of the student had a negligible effect on student performance. Furthermore, the fact that the gender of the student did not influence student performance is good news especially from Sub-Saharan Africa and shows that, with enough resources, there is no difference in academic achievement between male and female students.
This research has made significant contributions to the field of educational research, particularly in understanding the factors influencing student academic performance and the impact of school groupings on such success. Identifying the root of the issue is an essential first step toward implementing effective and enduring solutions. With the Kenyan government ambition to ensure 100% primary to secondary school transition, coupled with the newly introduced competency-based curriculum, it is crucial that students and schools are provided with the necessary resources to enable them to succeed in secondary school (Kenya Institute of Curriculum Development). As shown in the frequency Table, many secondary schools in Nyanza and Kenya are small day-only, mixed schools. Due to the small size of these schools, the government often overlooks them and instead focuses on the big, boarding, national, and county schools, which have undergone rapid expansion in recent years. These large boarding schools are out of reach for most parents because they are expensive. One approach to improving performance in these small schools would be to consolidate them to optimize resource allocation.
One limitation of this study is that the cut-off chosen was a C + and it’s not known if a different cut-off would produce a different classification accuracy. Secondly, data is based on the 2011 KCSE results so a newer data set can be used to check if the results hold. Thirdly, admission to secondary schools is determined by students' performance in the Kenya Certificate of Primary Education (KCPE), with high achievers securing places in large, boarding, single-gender institutions. Therefore, it is crucial to consider KCPE performance as a predictor for future studies. Additionally, in schools that provide both day and boarding options, differentiating between day students and boarders was not feasible due to data limitations.