Statistical modelling
From the results of the two analyses we can break down the
sum of squares of 445.076 for DAMAGE7 in the first analysis into components for
DL and DQ (101.581 and 308.044 in the second analysis - see previous screen
) and a remainder (445.076 - 101.581 - 308.044 = 35.451). Presenting these
values together with the residual line we get:
|
Source of variation |
d.f |
s.s |
m.s. |
v.r. |
|
DAMAGE7 |
6 |
445.076 |
74.179 |
|
DL |
1 |
101.581 |
101.581 |
|
DQ |
1 |
308.044 |
308.044 |
|
Remainder |
4 |
35.451 |
8.863 |
1.80 |
Residual |
683 |
3357.495 |
4.916 |
|
|
The 'Remainder' term, which represents the DAMAGE7
variation not accounted for by the quadratic function, is not significant (VR =
1.80). Since the size of this remaining variation is not statistically
significant it can be deduced that the quadratic fit is a good one. We can also argue that it
is not necessary to add a cubic term to the polynomial equation and decide not
to do so.
|