Multivariate Analysis of Performance of Secondary Schools in Simiyu Region, Tanzania

Ramkumar Balan

Background and Objective: The candidate’s skill and mastery of competencies by the grades highly depend on the school’s performance. No studies were yet found in Tanzania dealing with the 5 significant variables together in a multivariate vector approach even though studies separately considering some of the variables were found. The general objective of the study was to cluster the secondary schools of Simiyu Region of Tanzania according to their performance. Materials and Methods: The study is based on data available from 44 secondary schools selected from the Simiyu Region of Tanzania. As 34 Government schools and 10 private schools were involved and 18/28 urban schools and 26/133 rural schools were considered. The performance of schools was analyzed by applying multivariate tools, descriptive statistics, discriminant function and Mahalanobis distance. The performance of schools considered five variables-Number of students who sat for the examination (Sat), number of students who passed the examination (Pass), Grade Point Average (GPA), regional ranking (RR) and national ranking (NR) of schools. Results: Linear discriminant function 1 is more appropriate to classify the schools as per their performance. Stacked histogram 1 is efficient to perform the disjointing of three groups of schools without duplication or omission. The GPA and RR are the most influencing factors leading to the performance of schools. Most of the private schools were classified as high performing while 2/3rd of government schools were lacking in moderate performance. The number of students enrolled and sitting in the examination affects the rate of performance. Conclusion: The GPA contributes a higher part to the assessment of performance. Among the 44 schools under consideration, 14 were classified as high performing. The authorities and education institutes can adopt strategies to improve GPA and RR to secure high rankings in performance.

View Fulltext Back