Mathematics

Permanent URI for this collection

https://hdl.handle.net/20.500.11888/3206

Browse

Browsing Mathematics by Author "Dr. Ali Barakat"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Estimation of Multivariate Multiple Linear Regression Models and Applications
    (2015) Jenan Nasha’t Sa’eed Kewan; Dr. Mohammad Ass’ad; Dr. Ali Barakat
    Regression is a statistical method which is widely used in various fields of science for defining the relationships between the variables in the form of an equation to estimate the parameters, the strength and direction of the relationships. This main objective of this thesis was to study the Multivariate Multiple Linear Regression Models which relate more than one dependent variable with more than one independent variable. Four types of regression models were considered; Simple Regression, Multiple Regression, Multivariate Regression, and Multivariate Multiple Regression. The method of least squares was used in estimating the multivariate multiple linear regression models. Then multivariate analysis of variance (MANOVA) was used to test the usefulness of the estimated models. Several software programs were used to achieve this objective, such as Stata, Matlab, Minitab, SPSS, and SAS. The multivariate multiple regression model was applied to simulated data and to real data. A case study was constructed from three universities in Palestine; An-Najah National University, Arab American University, and Alquds Open University. A sample of size 350 students from these universities was considered to study the relationship between two psychological variables (self concept and achievement motivation) and the cumulative average of the student, which were the response variables. Then we showed how these three responses were affected by some predictors such as tawjihi average, English score level exam, budget per day, absents per semester, the study program of the student, and the university he/she is studying in. The results showed, according to the p-values, that the study program has no effect on both self concept and achievement motivation of a student. When we test the effect of the study program on these two responses, the p-values were 0.5674, and 0.2227, respectively. On the other hand, the cumulative average was affected by the study program, the p-value was 0.0501. The cumulative average of the student was also affected by the university, it was with null p-value. But the student’s self concept and achievement motivation were not affected by the university, the p-values were 0.2530, 0.4352, respectively. Also, data for each university was analyzed separately. The results for An-Najah National University showed that the twajihi average controlled the responses very well and the student’s cumulative average was affected by the study program. The results of The Arab American University showed that the three responses were affected by tawjihi average and English score, but they were not affected by the study program. Finally, the results of Alquds Open University showed that the three responses were affected by just one factor which is the tawjihi average.