Statistical modelling
With the codes DL and DQ representing the linear and
quadratic terms for age of dam, respectively, the output, as shown, can be
obtained.
We see that the residual mean square is slightly increased
from the value of 4.916 kg2 in the previous model to 4.939
kg2 here, implying a slightly poorer fit.
Note, however, that the sum
of squares accounted for by each of the other factors is unchanged as they
occupy the same positions in this analysis as they did in the previous one and
are fitted before DL and DQ.
However, let us look at the analysis a little more
closely.
|
|
***** Regression Analysis *****
Response variate: WEANWT
Fitted terms: Constant + GENOTYPE + YEAR +
SEX + AGEWEAN + DL + DQ
*** Estimates of parameters ***
Parameter |
estimate |
s.e. |
t(687) |
Constant |
2.702 |
0.929 |
2.91 |
GENOTYPE RD |
-0.389 |
0.220 |
-1.77 |
GENOTYPE RR |
-1.040 |
0.271 |
-3.84 |
GENOTYPE DR |
-0.511 |
0.304 |
-1.68 |
YEAR 92 |
-1.565 |
0.293 |
-5.34 |
YEAR 93 |
-1.099 |
0.276 |
-3.99 |
YEAR 94 |
-2.820 |
0.359 |
-7.85 |
YEAR 95 |
-3.215 |
0.346 |
-9.30 |
YEAR 96 |
-2.342 |
0.391 |
-5.99 |
SEX M |
0.476 |
0.170 |
2.81 |
AGEWEAN |
0.07026 |
0.00886 |
7.93 |
DL |
2.188 |
0.249 |
8.79 |
DQ |
-0.2688 |
0.0340 |
-7.90
|
Change |
d.f. |
s.s. |
m.s. |
v.r. |
+ GENOTYPE |
3 |
570.427 |
190.142 |
38.50 |
+ YEAR |
5 |
735.646 |
147.129 |
29.79 |
+ SEX |
1 |
59.013 |
59.013 |
11.95 |
+ AGEWEAN |
1 |
336.792 |
336.792 |
68.19 |
+ DL |
1 |
101.581 |
101.581 |
20.57 |
+ DQ |
1 |
308.044 |
308.044 |
62.37 |
Residual |
687 |
3392.947 |
4.939 |
|
Total |
699 |
5504.450 |
7.875 |
|
|
|