In: Statistics and Probability
Joglekar, Schuenemeyer, and LaRiccia (1989) looked at the
relationship be-
tween the percentage of hardwood in a batch of pulp and the tensile
strength of the
paper. The question is can the hardwood percent useful in
predicting the strength of
the paper.
psi | % |
6.3 | 1 |
11.1 | 1.5 |
20 | 2 |
24 | 3 |
26.1 | 4 |
30 | 4.5 |
33.8 | 5 |
34 | 5.5 |
38.1 | 6 |
39.9 | 6.5 |
42 | 7 |
46.1 | 8 |
53.1 | 9 |
52 | 10 |
52.5 | 11 |
48 | 12 |
42.8 | 13 |
27.8 | 14 |
21.9 | 15 |
(a) (2 points) Create an ANOVA table using only the single
predictor variable hardwood.
(b) (2 points) Compute and interpret the value of R2.
(c) (2 points) Give the hypothesis and conclusion of your ANOVA
table (Signicancer is 5%).
(d) (2 points) Find the RMSE and interpret this value.
(e) (4 points) Add a third column to the data set by squaring the
values in the %
hardwood column to investigate a multivariate model y^ =
b0+b1x+b2x^2+E". Create
an ANOVA table using both the variable hardwood and the variable
hardwood2.
Calculate and interpret R2.
(f) Provide some graphical evidence AND explain the dramatic change
in
R2