Question: (30 points total) Regression .

1. Construct a model to predict weight from elbow diameter. Fully report the model as practiced in the quizzes and as shown in Cronk. Test the assumptions of the model, and state your findings and conclusions. What is the residual for subject number 50? ( 10 points )
2. Create a regression model to predict weight based on any or all of the variables other than weight variables (so you should start with a total of 16 predictors in the model).
   1. Use what you know of model reduction to reduce the model. Remember, the purpose of the model is to be able to predict people’s weight. Use your (un)common sense, and all the principles relevant to model reduction, to arrive at your best choice of regression model to predict weight. To assist you in your search, a copy of the correlation matrix has been added to the last page of this midterm. Your grade on this question will be determined more by your explanation of your choices at each stage of model reduction , than by the final model you end up with. So make sure you include a good argument for your choices at each stage of the process . Also include the R ² , ANOVA, and coefficient tables from the output at each stage of the analysis. Once you have arrived at your final model, fully report this model, using the format laid out in class and in Cronk. ( 20 points ). What is the residual for subject number 50 in this model?

Price: $2.99

Solution: The downloadable solution consists of 11 pages
Deliverable: Word Document