Isds Case
ISDS CASE 16.2 2012
Executive Summary:
We believe that utilizing Regression Analysis will provide us with enough data to make an accurate prediction. Regression Analysis is simply put using the value of one dependent variable Y based on the data of other independent variable(s) X. To conduct a Regression Analysis, we need to input data from two variables and produce a Regression equation that describes the relationship between the dependent variable and the independent variable. A dependent variable is the variable to be forecast, i.e. what we want to predict. An independent variable is what we believe the dependent variable is based on.
In this case, we are trying to determine the best regression model to predict a student’s University GPA (Y, the dependent variable) based on the grades they received in high school (X, the independent variable). There are currently two options for the Y, independent variable, and this result is made possible by two methods (Xs) to predict a student’s University GPA (Y):
1. Method 1 (X1): utilizes the student’s average from the BEST 6 courses taken in High School. 2. Method 2 (X2): utilizes the student’s average of English and Calculus, plus the Best 4 courses taken in High School.
The goal is to change University admission standards in accordance with the best Regression model utilizing either Method 1 or Method 2. To find the most efficient model, which predicts a student’s University GPA best, we must analyze two regression models based on each of the methods.
When comparing the values we can conclude the BEST 4+EC model is a more effective model to predict University GPA. The B4+EC model is 35% accurate. The B6 model is only 24% accurate. The R2 value (coefficient of determination) shows us the strength of the relationship between x and y. This relationship was stronger in the B4+EC model by .113 GPA units. This translates to approx. 11% more accuracy when compared to the B6 Model. In addition the
Executive Summary:
We believe that utilizing Regression Analysis will provide us with enough data to make an accurate prediction. Regression Analysis is simply put using the value of one dependent variable Y based on the data of other independent variable(s) X. To conduct a Regression Analysis, we need to input data from two variables and produce a Regression equation that describes the relationship between the dependent variable and the independent variable. A dependent variable is the variable to be forecast, i.e. what we want to predict. An independent variable is what we believe the dependent variable is based on.
In this case, we are trying to determine the best regression model to predict a student’s University GPA (Y, the dependent variable) based on the grades they received in high school (X, the independent variable). There are currently two options for the Y, independent variable, and this result is made possible by two methods (Xs) to predict a student’s University GPA (Y):
1. Method 1 (X1): utilizes the student’s average from the BEST 6 courses taken in High School. 2. Method 2 (X2): utilizes the student’s average of English and Calculus, plus the Best 4 courses taken in High School.
The goal is to change University admission standards in accordance with the best Regression model utilizing either Method 1 or Method 2. To find the most efficient model, which predicts a student’s University GPA best, we must analyze two regression models based on each of the methods.
When comparing the values we can conclude the BEST 4+EC model is a more effective model to predict University GPA. The B4+EC model is 35% accurate. The B6 model is only 24% accurate. The R2 value (coefficient of determination) shows us the strength of the relationship between x and y. This relationship was stronger in the B4+EC model by .113 GPA units. This translates to approx. 11% more accuracy when compared to the B6 Model. In addition the