Abstract

MULTIPLE LINEAR REGRESSION MODEL: A STATISTICAL TOOL FOR PREDICTION OF SCORES OF FINAL YEAR MATHEMATICS DEGREE STUDENTS OF THE COLLEGE OF EDUCATION, AGBOR IN AFFILIATION WITH DELTA STATE UNIVERSITY, ABRAKA

Ohoriemu Blessing Okeoghene, Osemeke Reuben Friday and Aghamie Sunday Osiebuni

---

Multiple linear regression is one of the most widely used statistical tools in diagnosing the performance of students in any examination. It is defined as a multivariate technique for assessing the correlation between a dependent variable(Y) and some combination of two or more independent variables(X1, X2, X3 . . Xp). In this paper, a multiple linear regression model comprises of three independent variables(X1, X2, X3) is developed to analyses the performance of final year mathematics Degree students of the College of Education, Agbor in affiliation with Delta State University, Abraka. The model is based on the data of student?s scores in first tests, second test, and class attendance. The estimates both of the magnitude and statistical significance of relationships between the variables have been provided. Several statistical measures such as descriptive statistics, F calculated, T calculated, coefficient of determination(r2), adjusted coefficient of determination, Mallows Cp Statistic, multicollinearity diagnostics and graphical residual plots, were used as a benchmark for selection of best subsets optimal regression models in a multiple regression diagnostics and a statistical tool for the analyzing the performance of final year mathematics Degree students of the College of Education, Agbor in affiliation with Delta State University, Abraka

PDF