Bakker Daniel. K.2024-09-052024-09-052019-092455-7838(Online)https://ir.ueab.ac.ke/handle/123456789/16Generally, the University students have a vast variety of subjects in the core area which is sometimes problematic for the students to comprehend. They have chosen social media or rather internet, where there is a lot of learning materials concerning the area of their specialty. Predicting the academic performance of the students helps the instructor to develop the virtuous understanding of student’s community and take the fit procedures to make their learning comfortable. MLR was used to build a model which predicts the performance of the students in the discipline of Statistics. Based on the collected data, the predictors of this model focused on how many hours spent on the social media. The predicted Cumulative GPA at every academic semester period was the dependent variable. The results revealed that the predicted model scores gives a better accuracy with M 1 through M 3 in the range of 5%  from the original scores. Therefore, the instructors can analyze the performance of the students and can also extemporize the teaching techniques used based on the result, of not only Statistics students, but also other areas they handle.en-USArticle