This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Computer II project thoughts-common errors For EVERY final model that you find, make sure you check there 4 assumptions, make the predictions and confidence intervals. Make sure you have read the FAQ-p. 9. Use the units the data are presented in unless you are otherwise instructed. If you do an additional analysis finding the best model with the “order” variable and all other independent variables, make sure that you include in your Executive Summary the best model without day, standard deviation of “y” scores about that model, R^2 and any prediction and confidence results in a paragraph, followed by another paragraph with the best model with day and similar results. Note in each of these paragraphs if you have any concerns about the four assumptions that have been noted in the analysis. Using Principle 4, check if terms can be dropped-which means do a test for each eligible beta if Ho: beta=0 vs H1: not Ho In checking assumption 4, include the result of your inspection of the e(t) vs e(t-1) plot (guided by the “y-bar”...
View Full Document
- Fall '10
- Regression Analysis, Best Model, Final Model