Using a dataset of your choice, estimate the regression equation and implement it in a spreadsheet model. Use the process model described below to perform your analyses and produce your report in Powerpoint. Please attach your excel file(s) too.
Procedure of Building Regression Models
Define Objectives
Define/Clarify Purpose. Explain the scenario briefly. What relationships are you trying to find? What is the dependent variable? What are the independent variables?
Select Variables
Identify possible independent variables (predictors). Use scatter plots and correlations for selection. You can also use different types of descriptive analysis:
- Univariate – draw frequency histograms of each variable. Compute means and standard deviations of each.
- Bivariate - plot each independent variable individually against the dependent (scatter plot) to see if it seems to be related to it.
Perform Modeling
Perform multiple regression analysis. Drop any insignificant variables, run regression again. Final model must have variables that are all significant.
Provide Interpretation
Interpret the results. Write the mathematical model; explain what it means, interpret the R-squared value and the significance of the F value. Discuss what other variables may be relevant that were not studied, which may improve the model in the future.
Is possible, compare predictions and actual values. Modify/Refine/Expand model is necessary.