Dissertation (MSc Statistics)This study has been conducted with the main objectives to investigate the pattern of students’ performance in mathematics and to forecast the performance of students in mathematics in secondary schools in Zanzibar for next 6 years on the basis of best selected model. The study used secondary data of all 64,446 students from 123 schools of Zanzibar, who sat for final examination from 1986-2016. The yearly data of students' results were collected from Ministry of Education and Vocational Training Zanzibar ( MoEVTZ) and the yearly time series examination result data for CSEE in Zanzibar from 1986-2016 were compiled for Male students , Female students and total students combined in the form of percentage credit, percentage pass and percentage fail. These data were further analyzed for checking and testing of normality. Normality assumption was validated for male students, female students and total students individually by using Kolmogorov-Smirnov test and Shapiro-Wilk’s test. Further, in this study, various Autoregressive Integrated Moving Average (ARIMA) model have been fitted to see the pattern of performance of male students, female students and total students. The best fitted model has been selected on the basis of Akaike Information criterion and Bayesian Information criterion. ACFs and PACFs have also used and the best ARIMA models have been identified as ARIMA (1, 1, 2), ARIMA (2, 1, 0) and ARIMA (1, 1, 0) for male students' percentage credit, percentage pass and percentage fail, ARIMA (1, 1, 1), ARIMA (2, 1, 1) and ARIMA(2, 1, 2) for female students' percentage credit, percentage pass and percentage fail and ARIMA(2, 1, 0), ARIMA (1, 1,2) and ARIMA (1, 1, 0) for total students' percentage credit, percentage pass and percentage fail respectively. Accordingly, the forecast for next six years for all the three cases have been presented.