# Lesson 10

1. Cost of materials is determined most by New Capital Expenditures followed by Number of Employees. Overall a low R^2 value indicates nothing is a strong predictor of the cost of materials. 2. Dividend per share is the greatest indicator of average P/E ratio followed by average yield and earnings per share which each have a negative correlation. in general this is a very weak model with an R^2 of 0.18. 3. This model is very strong compared to the previous ones, however, no one independent variable takes a strong lead in predicting the outcome of NIKKIE. 4. This is a very strong model with Houshold income being the major predicting variable for food spending. My model indicated a R^2 of 1 and 0 coefficient for each variable. I don’t think this is possible though my methods were correct. # Lesson 9

13.2 13.6 The coefficients indicate that dividend payout is the more determinate of the insider ownership than is the debt ratio.

# Lesson 8

15.2 15.4 The errors indicate that the model drastically under predicts the number of acres of harvested tomatoes. New model is my recommendation.

15.10 15.12 S&P Index # Lesson 7

15.10 Both Models fit the data very closely given new orders appear in a nearly linear increase over the selected years.

15.11 The model has a relatively strong R^2 value indicating its decent strength at predicting values of the model. As the model progresses it appears to break down in accuracy of the data as unions become a socialist thing of the past and membership become sporadic.

15.12 The Linear model is useless for these data given the trend reverses direction during the 6th year. The quadratic form is much more accurate for this purpose as it reflects this change in direction. R^2 is much stronger for the quadratic model.

S&P

A fourth order model matches the data slightly better than a third order and much much better than a quadratic equation. Predicting the market is as profitable as taking the rablin express to cripple creek, not at all. 