APPLICATION OF CANONICAL CORRELATION ANALYSIS ON SCIENCE PRODUCTION

The development of any given nation is behind its advancement in science and technology; this can be achieved through the upbringing of the new generation on the knowledge related to science and technology by putting in all efforts, factors and mechanisms that will easily aid the better understanding and interest in science and technology. This research work tends to investigate the production of science in some selected schools in Gombe state, Nigeria; that offer science as their core subjects. Three factors were used; School output [i.e. the grades obtained in science subjects(Mathematics(MTH), Physics(PHY), Chemistry(CHM) and Biology(BIO)], the School input [i.e. Averagely Equipped Library and Laboratory for Science (AELAL), Science teachers' years of teaching experience (STYTE), Instructional Hours on Science subjects per week (INSHR) and students' teacher ratio (STR)] and Environmental input [i.e. The number of text books on science possessed by students (NTBS), hour spent studying science outside school hours (HRSS), home leaning aids on science such as computer, science dictionary est. (HLAS) and home extra moral teacher on science(HETS)]. Two sets were formed, Set-A (school output) and Set-B (school input and environmental input).The data used is obtained through the questionnaire distributed to the random selected school. The research work adopts the use of Descriptive statistics to verify the normality of the data and Canonical Correlation Analysis to investigate the relationship between the sets of the data. Three Canonical roots were obtained and only two are statistically significant, the first showing a strong positive correlation coefficient between the sets of data, indicating the impact of the School and Environmental inputs on the school output. However, improvement on the School and Environmental inputs will equally improve the production of Science in the selected schools as a case study and some other schools in the states at large.