This study assesses whether five government and two non-government inputs into public high schools matter in the schools' performance in the National Achievement Test for fourth-year students. The government inputs are: the average number of students per teacher, the average number of students per class, the average number of students per room (counting laboratory, computer, home economics and other special rooms), the average number of students using a classroom seat, and the average number of students using a classroom seat, and the average number of students per toilet bowl. The non-government inputs were the drop-out rate and the total enrolment of the school. The results showed that both non-government variables were relevant to performance on the test. Of the five government variables, only two mattered: the average number of students when all types of rooms were counted, and the average number of students per toilet bowl. It was also found that the seven variables could account only for 2.5% of the total variability in the test, leaving 97.5% unaccounted for. Suggestions were given in order that more of the remaining 97.5% variance could be explained in the future.
