Assignment Objective: Complete a full regression analysis of your data.
Regression analysis leads to a model or equation that predicts the value of the
dependent variable. Independent variables are the inputs for the model, and the
dependent variable is the output.
This analysis should use all the variables in your data set to create a multivariate model.
As part of the analysis, some of the variables may be removed from the model. There is
no maximum or minimum number of variables in the final model. All steps of modeling,
including removal of variables, need to be well documented.
Procedure:
1. Conduct regression analysis from start to finish.
2. Start by running regression.
3. State the initial regression model and the model strength.
4. Graphically check the regression output to see if any violations of regression
requirements (assumptions) occur. Consider transformations if needed.
5. Remove predictors from the model sequentially if needed.
6. Determine the best overall model.
7. Show the final model.
8. Indicate the strength of the model.
9. Make business suggestions. (I have done most of this part) (Feel free to add to it or change it if its easier for you to write it using a different model)
The written portion of this assignment should start with the initial model, cover at a high
level the steps to refine the model, detail the final model, and cover at a high level the
reliability of the model that has been created. A general idea of how the model can be
used in a business context should be included as well. Two popular contexts are to use
past data to predict future values and to check for over-/under-predicted values.