Using Multiple Factor Analysis to Study Students’ GPA: Case Study Faculty of Science, University of Benghazi.

Abstract

Data or dimensionality reduction is very common and critical in statistics because, in the model building phase in any statistical research or study, the researchers are always annoyed by the number of variables to be included or excluded in a model. In this sense, Multiple Factor Analysis (MFA) can be employed to select variables and reduce data. In other words, it can be used to know the variables that have the largest amount of variation in the data understudying. The aim of this paper is to study the students learning behavior to find out the variables that have the largest amount of variation by applying the Multiple Factor Analysis (MFA) to the students' GPAs in different departments in the Faculty of Science-University of Benghazi and to find out whether the variables (courses), in the database, have a considerable variation that worth studying and investigating. The variables considered in this study are the different courses in each department and the criteria used to measure these courses. The data collected in this study also contains information related to the course subjects and grading systems in these departments. The results of this study show that fifteen variables (Courses) absorb most of the variation in the data sets. The finally selected fifteen variables vary in the amount of the distributed variation. The results also show that the variables used in this study correspond to some of the variables and data reduction used in Othman and Gebril, 2014, using Block Principle Component Analysis (BPCA). It is hoped that the findings of this study will be a useful contribution to our faculty to improve teaching methodology and provide these different departments with modern educational facilities.